Few days back while searching some electronic components on Ebay I stumbled upon a link from some Russian guy selling vintage Soviet era made IN-12B Nixie tubes. I always wanted to use Nixie tubes in my projects so I ordered them 12 of them from that seller. The tubes I got were probably made in August 1984 (that’s what written on them). These Nixie tubes were used to display digits in frequency counters, military clocks and even calculators before 7-segments became mainstream. The particular thing I like about them is their beautiful orange glow in the dark. Now these tubes are out of production and as of my knowledge only one person makes them in the whole world. Let’s discuss more about them.
Nixie tubes are vacuum sealed glass tubes similar to vacuum tubes and Neon lamps and are filled with a mixture of Neon, Mercury and Argon. The anode in a Nixie tube is a mesh of wire on the top and cathodes/digits are pieces of metal in the particular shape of that digit; stacked and separated by ceramic connectors. Here is how mine IN-12B looks like
In an ordinary neon lamp, the metal cathode is heated up to a relatively high temperature (sometimes well over 100°C so negatively charged electrons “boil” off it and race toward the positively charged anode, colliding with gas molecules in the process, exciting them, and making them give off coloured light (as we explain fully in our article on neon lamps). In a Nixie tube, the cathodes remain relatively cool and that’s why they are described as “cold cathode tubes”. But the gas mixture that surrounds them is at a very low pressure (perhaps 1/100th of normal atmospheric pressure or even less). When a voltage of about 200 volts is applied between the anode and one of the cathodes, the low-pressure gas becomes ionized (its atoms or molecules are turned into positively charged ions and negatively charged electrons). When the ions and electrons collide with metal , the metal atoms are ejected from the cathode and we get a glowing orange colour all around the cathode that follows its shape. Effectively making it appear as though one of the numbers 0–9 is illuminated.
One thing to note about these tubes is that they require a relatively high DC voltage of around 170-200V to operate. But they don’t require too much current, just 2-3 mA per digit. I didn’t have 200V-30mA DC power supply, so I had to make one.
There could be two ways to make a 200V DC power supply. One is to rectify the 220V rms mains AC voltage and then limit the current by some resistor network, and the second is to boost a low DC volt supply to 200V. I found the 2nd option to be little safer, so I used a 555 timer to switch current through an inductor via a high voltage MOSFET as seen in the schematics.
The 555 is configured as an a-stable oscillator which oscillates at a frequency of about 31kHz determined by resistors R1 and R2, and capacitor C2. The 555s output directly drives a high voltage MOSFET IRF740, Q2, that switches the current through an inductor, L1. When the FET is on, current flows from V+ through the inductor to ground. As the current in the inductor builds up, the FET then turns off. When the FET turns off, the current flowing through the inductor tries to continue to flow but can’t flow through the FET, so the voltage on the FET’s drain terminal rises until the ultrafast diode, D1 becomes forward biased. This allows the energy contained in the coil to be dumped into the high voltage capacitor, C4.
This cycle continues until the voltage across C4 reaches a value set by potentiometer RV1. R4, R5 and RV1 form the feedback divider for the circuit, and are chosen such that the voltage across C4 is divided down to around 0.7 volts. This voltage is applied to transistor Q1, a BC547. When the voltage across C4 rises to the point where the voltage from the divider is enough to turn Q1 on, then Q1 pulls down the Control pin of the 555, stopping it from oscillating and shutting down the converter.
As soon as this happens, the voltage across C4 begins to fall, as does the voltage into the base of Q1, and so Q1 turns off, allowing the converter to restart. This is how the circuit regulates the voltage, and in practice it works quite well. Resistor R3 and capacitor C3 form a simple snubber network to suppress the voltage transients around the drain and source of the FET, while C1 is the main reservoir cap for the circuit.
Putting it all together: