Flybacks and Plasma Tubes

[alan sailer]

I'm building another HV AC power supply for powering up homemade plasma tubes and a question has come up. It's not important,
more curiosity.

The plasma tubes are powered by AC because there is no internal electrode. I wrap a piece of adhesive copper tape around the base of the tube (area of about 60 cm^2) and use spring fingers to contact this electrode. As I understand it the foil acts as a series capacitance to the output of the flyback, conducting the current into the tube to create the plasma.

Recently I did a few simple experiments to see if this electrode could be smaller and still function well. The answer is yes, with a 12 cm^2 electrode things work quite well. Even smaller still works.

The question is what would be an optimal electrode? I'm not even sure what I mean by optimal. Best match to the flyback transformer?
I have seen some papers from Information Unlimited where the writer talks about matching different capacitive loads to a flyback but I never understood what he was talking about.

The circuit I am using to power the flyback is an old PVM400 from Information Unlimited. I got a bunch of these supplies from a plasma tube artist because they were broken. In most cases the output flyback was fried and I use 70's vintage color TV flybacks as a new output transformer. Another issue that I've never explored is optimizing the new flyback to the PVM400 circuit. The existing primary has always worked fine so I have not bothered to go any deeper.

[davekni]

The question is what would be an optimal electrode?

If you know your glass thickness and approximate dielectric constant (likely around 5), calculate capacitance from electrode to internal plasma gas.  If that capacitance is well larger than capacitance from globe to space, globe-to-space capacitance will dominate net series capacitance.  Sensitivity to electrode size should be small.  Once net capacitance changes, all depends on flyback and driver circuit.  Flyback secondary has winding capacitance and load will behave as a lossy capacitance.  Perhaps "optimum" is when total capacitance resonates with secondary inductance at drive frequency.  Or perhaps primary drive is also resonant, making a dual-resonant circuit.

[alan sailer]

David,

Makes sense. Is there any way to ballpark the external, ie plasma to environment, capacitance? Since I don't live in a metal box I'd imagine it's pretty small.

[davekni]

Makes sense. Is there any way to ballpark the external, ie plasma to environment, capacitance? Since I don't live in a metal box I'd imagine it's pretty small.

I'd just make a guesstimate of what size sphere would be like your plasma container.  Capacitance of a sphere to infinity is radius_in_meters * 4 * PI * 8.854 pF/meter.  111pF for a 1m radius, so 11.1pF for a 10cm radius (20cm diameter).  Would be slightly higher given grounded objects aren't infinitely far away.

[alan sailer]

David,

Cool. I'll run a few numbers.