Plasma Striations Demonstrations

When lighting fluorescent tubes with induction (Tesla) coils, I often have seen plasma striations, patterns of light and dark, often flowing along portions of the tube, These patterns are occasionally seen in conventionally powered fluorescent tubes, usually in cold weather or end-of-life conditions. Striations are also reported in fluorescent being run off of a dimmer. The consensus is that these striations are an artifact of low current density in the tube under observation.

How often do you get to work with ionized low pressure (2.5 Torr) argon, ultra-violet light, RF coils and high voltage for under $100?

[In the images below, the playback rate is too high by a ratio of 30fps/24fps.]

Striations in F15T8/CW Tube Driven by Induction Coil

Download Sample 3 Second Cool White Striations Download Full Resolution Cool White Striations

Hardware Notes

This investigation uses two Tesla coils:
  1. The ‘Acrylic’ made of 434 turns of 30 gauge wire (200 feet) wound on an acrylic tube 1.758 inch in diameter. This coil measures 23 Ohms. 2.63mH and is  driven with a 28Vpp sinewave around 1.628 MHz.
  2. The ‘PVC’ is  made from 585 turns of 30 gauge wire (also 200 feet) on a 1 inch PVC form. This coil measures 23 Ohms. 2.63mH and is driven with a 28Vpp sinewave around 1.742 MHz.

The bottom of the winding is driven by a function generator at or just below resonance, while the top is open. By adjusting the frequency, I can get the tubes to light automatically when power is present.

However, for greater insight into the physics of the coil and plasma interactions, I prefer to use a setup that requires human interaction to start and latch the light. I set the frequency of the function generator to be about 3% below resonance. To strike the arc in the tube, I move my hand near the tube. My body capacitance lowers the resonant frequency of the coil. At the proper separation, resonance is achieved. The resulting high voltage ionizes the tube, resulting in a transformer coupling to the plasma, as well as the plasma providing the second plate of a capacitor (the coil windings being the other), keeping the circuit at resonance as my hand is withdrawn.

Initially, the tube was hand-held near the coil. Other variations included a T mount, initially using a pasteboard cradle seen below, later using a groove ground in the top of the form, and vertical mounts, sometimes using the plastic tube of the winding, other times using a spent spool former as a base.

Experimental Work

My initial work used a F15T8/CW tube. This is about 15 inches long, and is a convienent length to use for public displays of Tesla coils. This also happens to be a common desk lamp tube, as noticed in the some of the background pictures. This tube has a clamping breakdown voltage around 55V, and is usually run at 300 mA current.  These are low pressure tubes, about 2.5 torr for the Argon, with mercury partial pressures much lower, around 50 mTorr.  I am driving these coils with RF level signals around 1.7 MHz and voltages on the order of 1-2kV. Electrical wavelengths (lambda = c/f) at these frequencies correspond to 170 meters or so, Consequently, the striations are likely not due to standing waves in the RF excitation. (They may be related to the plasma frequency of the mercury vapor.) I wanted to see the striations better, and the phosphour coating was in the way.  So, I changed to germicidal tubes, G15T8, which uses a quartz tube (transparent to UV) and no phosphour. This is a definite eye hazard. I recommend using video recordings of the plasma glow, rather than repeated optical observations with your last two good eyes. However, the clear tube lets us see important details.

Striations in G15T8 Tube Driven by Induction Coil

Important Observations

Details of Plasma Generation Zone

Next are series of vertical mounted tubes. We see the neutral flow point corresponds to a point about 2/3 up the winding for the closest tube/winding position. We also  see some downward flow in the furthest positions.

Vertical Closest to Coil

Vertical Further Out

Vertical Still Further Out

Vertical Furthest Out, Some Downward Flow


The electric field to drive a fluorescent tube is, by design, very low at 55V/15inch, 150 V/m, or 1.5V/cm. The mean free path for Argon (71pm radius assumed) at pressures of 2.5T is quite small compared to the tube, at  0.14 mm.  The excitation energy (but not ionization) for Argon is about 11V. The excited Argon efficiently transfers energy to the mercury (10.4V) which ionizes. Upon recombination, the mercury radiates with the bulk of the energy concentrated in UV around 254 nm and 185 nm. Clearly multiple collisions are required to get electrons up to activation energy for the Argon. I am having confusion reconciling the scale and location of the ionized bubbles with my expected field around the coil. I am comfortable with field levels of 60V/mm, as witnessed by NE-2 bulb ionizations. I need to find a few more numbers, such as the carrier density, recombination time, and plasma frequency before I feel I fully understand this. I am also pondering the cathode fall regions of an electrode, versus electrodeless operation of the plasma here.

Links to Large MOV files