Nixie tube at 300 000 V
I had a few spare IN-18 tubes from the construction of my clock, in case some of them failed (which happened). I was thus left with both non-working and working nixies ... plus a Tesla coil. How the story goes is pretty obvious at this point :
Hyperion fires several amps through a working IN-18 nixie tube.
This "experiment", apart for producing very pretty results, highlight some aspects of nixie tube operation (in normal conditions). We will compare how a working and a fractured nixie react to this rough treatment. I suggest you read this short description of operating theory before going further.
Nixie tubes contain a very low-pressure gas mixture composed primarily of neon. When this medium is exposed to a high enough voltage, it will become conducting. In the process, electrons will be ripped apart from their atoms and recombine with other ones.
The main glow is caused by de-excitation of the gas mixture atoms while the yellow glow around the metallic parts (digits and rods at the tube's base) is due de-excitation of sputtered atoms instead. I'm unsure about why some digits are lit but not other, but my guess is that some paths tend to be favoured in a spike effect-like phenomenon. The last important feature is the filaments that extend past the tube ; we'll talk about them shortly. Let's also remark that nothing happens through the glass (except some refraction) : ionization occurs inside and outside the tube but not through the glass, as it is an excellent insulator.
An important (yet obvious) thing to have in mind is that the nixie lights up, not because it is wholly subject to a high voltage, but because there's a great voltage difference between the head and the bottom of the tube.
The results are somewhat less spectacular.
As the gas mixture was already long gone from this tube, it acted like a normal conductor, with sparks preferably jumping from the sharpest areas (again, spike effect). The fact that they jump only from the far end regarding to the top load has to do with the geometry of the electrical field : the voltage difference is the greatest between the tube base (closest to the top load) and the tube head. The external edge of the head, to be more precise (see picture at the end of this page).
Here, the main glow is gone and we're left with "common" sparks. The reason for that is because the pressure of the atmosphere is too high for an entire volume of gas to ionise. Instead, as the existing path of ionised air facilitates the grow of the channel, the air is ionised "by segments" (hence filiform sparks).
On the last picture, the tube was mounted with head first with tape, and one can see some fancy colours around this point. These are most probably due to the tape burning. There was indeed some melted plastic on the tube afterwards.
It is important to understand that the big glow inside the working tube and the sparks outside it are exactly the same phenomenon : excitation then de-excitation of atoms. The striking difference between the two is due to the nature of the medium.
Inside the working tube, the gas pressure is very low and the entire volume can be made conducting. On the contrary, outside the tube or in the broken tube situation, the gas pressure is much higher (atmospheric pressure) and the formation of conduction zone is much more constrained, and only filiform zones (i.e. sparks) can be created.
Again, this has to do with the electrical field being much more intense at the tip of the spark (which is conducting, like a metal) that elsewhere ; sparks adopt this filiform shape because it is "easier" (i.e. it demands less energy).
The working tube eventually survived. It suffered from an increased cathode poisoning on the "8" digit, which was already slightly attacked. Indeed, this digit wasn't lit (see previous pictures) and sputtered material from the brightly lit ones settled everywhere, aggravating existing cases of cathode poisoning.