Unstable cathode-glow and
the flickering flame-effect neon

All gas discharges tend to flicker. The reason is that the gas breakdown voltage is dependent on the local pressure, and hence also on the gas temperature (which is not in equilibrium within the bulb). The breakdown voltage follows Paschen's law, which states that the relationships between pressure and breakdown voltage and electrode-spacing and breakdown voltage are both bathtub curves, i.e., the breakdown voltage can increase with either an increase or a decrease in either parameter. The discharge produces heat, and this changes the pressure in the region of the current path.
     Recall that a cathode-glow lamp operating in the constant-voltage part of its charactaristic has only part of the cathode glowing (full coverage occurs only in the anomalous-conduction region, just prior to arc formation). Hence, in such a lamp, particularly if it is operated well away from full coverage, the glow spot has a tendency to move around. The effect can be particularly pronounced if the electrodes offer a wide choice of locations for paths of nearly-equal distance and electrode work-function.
     In order to enhance the effect, for the purpose of making candle-effect lamps (etc.), the uniformity of the cathode electron work-function can be improved by the use of special coatings. That is the subject of US patent no. 3234808 (1966), in which Philip Kayatt describes the use of a barium azide process to produce amorphous barium-coated cathodes with a trace of barium nitride. We should however, be a little sceptical of his claims, for reasons dicussed below. Kayatt states that the fill gas can be neon or helium, or a mixture of neon and helium or neon and hydrogen. Variations of gas composition and pressure, electrode spacing, and series resistance (for a given power-line voltage), all modify the flickering effect.

flickering flame-effect neon bulb
click on the image to see an mpeg movie

Shown above is a flame-effect neon with integral ballast resistor for 230 V AC mains operation. It uses close-spaced electrodes (ca. 1mm apart) to make the flickering display visible, i.e., so that the optimum distance for gas breakdown lies in the electron and ion acceleration paths (i.e., the field lines) that occur from the outside of one electrode to the outside of the other. Flickering is enhanced by using flat electrodes, so that there is a wide range of approximately equal path lengths. Notice however, in the photographs below, that the spacing has been increased slightly at the part of the electrode where the width is narrow. This gives the glow a tendency to start at the wider part of the electrode area and move towards the tip. The behaviour is cyclical, but with a high degree of randomness.

flame effect neon, electrodes

flame effect neon, electrode spacing
Flame-effect glow-lamp electrodes.

Electrode spacing increases towards the top in this example. This gives a rising-effect in the motion of the glow spot.

The black electrodes are indicative of an amorphous barium and barium nitride coating.

The flickering flickery-thing-effect voltage regulator?
Shown on the right is a Mullard VS70 neon voltage regulator, manufactured sometime between 1940 and 1945. It pre-dates Kayatt's patent (filed in 1962) by about 20 years; but not realising that it has yet to be invented, it flickers anyway. The electrodes, in the spirit of wartime austerity, are made from two sheets of dull-grey magnetic metal (i.e., uncoated iron).
     Kayatt's patent will have expired in 1986, but its existence does pose some intriguing questions. What follows is partly speculation, but is probably not far from the mark.
     The flickering of cathode-glow indicator lamps is generally to be regarded as a defect. This is particularly the case in safety-critical applications (such as aviation equipment), where a simple indicator in peripheral view should not draw unnecessary attention to itself. It is also likely, of course, that the effect will be considered generally to be annoying. Hence an attempt to find a solution, as is perhaps intended by a patent awarded to Ted Foulke in 1934 (US Pat. no. 1965588)

Flickering volatge regulator - click reload if the animation has stopped
Click to see MP4 movie
     Foulke, although he does not mention flickering, expresses an interest in obtaining uniform cathode illumination in glow lamps. He believes that the way to do this is to obtain uniform electron work-function over the whole electrode area, and describes a process involving barium and caesium azides. This cathode coating technique is no doubt associated with the Philips barium-azide process introduced in about 1925 for the manufacture of vacuum thermionic valves (and soon to be obsolete).
     Unfortunately however, Foulke's ideas on the criteria for uniform glow are flawed. The way to achieve it is to operate the lamp in the anomalous-conduction region, which means either using electrodes of small area or, if the area is large, pushing the tube somewhat close to the arc conduction point. With the tube operated at low current, the nice uniform work-function would have increased the probability of unstable cathode-glow. Also, the inclusion of caesium (normally used in photocells) is unnecessary in glow lamps and would have served mainly to blacken the bulb.
     Foulke's tubes were probaby very nasty and flickery when operated at low current, and rather short-lived. So, fast-forward to 1962, and Kayatt introduces a horrid flicker-effect novelty lamp, citing Foulke's patent as prior art, and using a barium azide process to enhance the unstable glow by making the electron work-function uniform over the entire cathode surface. It seems that; that which once produced a uniform glow now produces an unstable glow.
     The black barium cathodes in the flicker-effect neon certainly do help to ensure an unstable glow. That this technique was essential-enough to pay a royalty on it however is somewhat doubtful, as the flickering flickery-thing-effect voltage regulator attests. Also note, by taking a selection of pin-electrode mains-indicator neons (with integral series resistor), that every now and then, one will be found in which the cathode-glow coverage is incomplete. That means that the tube is not-quite working in the anomalous-conduction mode, for which reason the glow may be unstable, causing the lamp to flicker annoyingly.

US Pat. No. 1965588, T. E. Foulke, 1934. Electrode coatings for cathode glow lamps.
US Pat. No. 3238408, Philip Kayatt, 1966. Flicker glow lamps.
Bizarre sounds from a neon flicker-flame lamp (youTube video) Eric Wasatonic.



Discharge Tubes

Coil Resonance Experiments

© D. W. Knight 2013, © Andy Cowley M1EBV 2004.