Neon Tube

This build uses a small commercial indicator lamp, never a sign tube. Always limit the current with a series resistor of at least 100 kΩ, touch nothing while powered, and work with an adult. Never build a high-voltage tube yourself.

Claude says almost nothing about light. He talks about gas purity and electrode area. This is a patent about service life.

Put a neon lamp in series with a 100 kΩ resistor. Without it the lamp runs away and destroys itself in seconds.

Increase from zero while watching the meter. Nothing, nothing, nothing — then the lamp lights all at once. Record that value: it is the striking voltage.

Reduce gently. The lamp stays lit well below its striking voltage. An ionised gas conducts better than a neutral one — it takes more to start it than to keep it going.

Examine the lamp closely. The orange glow clings to one electrode only. This is not a filament heating up: the gas itself is emitting.

Light a small incandescent bulb beside it. One runs hot and radiates broadly, the other stays cool and gives a single colour. Two unrelated mechanisms.

Swap the leads. The glow moves to the other electrode. It always marks the cathode — which is why these lamps were used as polarity indicators.

On AC both electrodes glow: each is the cathode in turn. A sign works this way, and it is why both its ends wear evenly.

For 1 ampere the patent requires at least 1.5 dm², which is 150 cm². Draw that square: about 12 × 12 cm of electrode. This is why a sign has bulky end fittings.

The same current spread over a large surface gives a low density, so less local heating and less metal vaporised. You do not reduce the current, you spread it out.

Seal a scented cloth and some activated charcoal in a closed jar. An hour later the smell has gone: the charcoal has trapped the molecules. Claude does exactly this, but very cold.

Try again with charcoal straight from the freezer. The smell goes faster. Cold increases adsorption — and at liquid-air temperature the charcoal holds back almost everything except neon.

The patent. US 1,125,476, "System of illuminating by luminescent tubes", filed 9 November 1911 and granted 19 January 1915 to Georges Claude, a French engineer and industrialist.

This patent is not about neon, and not about light. That matters, because it is almost always summarised as "Claude invented neon", which is wrong twice over: neon was isolated in 1898 by Ramsay and Travers, and discharge tubes had existed for decades. What Claude contributes are two engineering fixes — purification by cold charcoal, and large electrode area — that take a tube from lasting days to lasting years. The patent's own wording is explicit: this is "the only case known" of a tube operating indefinitely without renewing the gas.

The neon came from his other business. Claude produced liquid air industrially, mainly for oxygen. Neon is a by-product of that distillation: present in air at about 18 parts per million, it had previously existed only in laboratory quantities. So Claude had both the raw material and the cold needed to purify it. It is a clear case of an invention arising from an industrial waste stream in the hands of the person looking for a use for it.

Why neon and not another gas. The patent says so: neon is barely absorbed by the electrodes. A tube filled with nitrogen or oxygen loses its gas into the metal and goes out; the noble gases react with nothing, so a single charge lasts. The orange-red colour is not an aesthetic choice, it is simply what neon emits — the other colours in signage come from other gases such as argon and mercury, or from fluorescent powders coated on the glass.

On the man, honestly. Georges Claude was a first-rank engineer — liquid air, neon, experiments in ocean thermal energy. He was also, during the Occupation, an avowed collaborator: convicted after the Liberation, imprisoned, and expelled from the Académie des sciences. Both facts are true and neither cancels the other. The patent remains a good patent; there is no reason to leave the rest unsaid.