Compendium of knowledge about Nixie

On a cold December morning in the Czech village of Březolupy, a man stops his truck in front of a 17th century castle. He puts on thick gloves, gets out of the truck and opens the tailgate. Then, very carefully, he unloads more crates of heavy equipment and glass.

That man is Dalibor Farny, who in 2012 decided to revive the production of Nixie lamps, the last commercial pieces of which were produced when he was still a child.

Also see:

Jak-pali-sie-1-1024x576
RBMK reactor power indicator (https://www.hbomax.com/pl/pl/series/urn:hbo:series:GXJvkMAU0JIG6gAEAAAIo)

These neon-filled fluorescent lamps were ubiquitous in the late 1950s and 1960s, finding their way into both domestic and industrial equipment. Their design originated in the basement of a German-American tinkerer in the 1930s, to be later commercialized by business equipment manufacturer Burroughs Corp. Nixie lights showed the data necessary for NASA’s moon landing, reported the current reactor power of nuclear power plants, and showed the ups and downs of stock prices on Wall Street. For many people, the warm glow of the Nixie evoked an era of unprecedented scientific and engineering achievements, exciting and tangible discoveries and seemingly limitless progress. Interestingly, it still does, even for people who, like Farny, grew up long after the lamps disappeared from common use.

In the 1970s, Nixies were overshadowed by LEDs, which were not only much cheaper to produce, but also more versatile. The lamps could have died a lonely death like countless other obsolete devices, but that didn’t happen. Their remarkable resurgence didn’t really begin until around 2000, when a small but dedicated group of hobbyists, collectors and enthusiasts began to seek out and buy old, never-used lamps and design their own clocks on them. Today in 2023, the movement continues to grow, and Nixies now occupy a unique niche in retro technology.

The Nixie lamp may seem like an unnecessarily complicated invention for something as simple as displaying a number. But if you think about what was technically possible in the mid-20th century, you can quickly conclude that it was the best possible decision.

Part of my Nixie lamp collection

Nixies operate based on a process called gas discharge. The phenomenon occurs when electrically charged particles, usually electrons, travel through the gas at high speed – about 2 percent of the speed of light. The accelerated particles collide with the atoms or molecules of the gas, ionizing them and creating an energetic plasma of charged ions and electrons. Ions excited to higher energy states give up excess energy in the form of photons of light. The color of the light depends on the gas: neon ions emit a red-orange light, hydrogen gives off a bluish-purple light, nitrogen gives off a spectacular violet, and krypton glows white-blue.

The earliest application of gas discharge was the Geissler tube, invented by German physicist Heinrich Geissler in 1857. The tube consisted of a simple glass bubble with electrical terminals at each end. The air removed from it was replaced with neon, argon or another gas. Some tubes were filled with conductive fluids, minerals or metals. Applying a direct current of several thousand volts ionized the gas, resulting in the emission of light. Scientists used Geissler tubes to better understand the nature of electricity, in addition to being sold simply as an ornament. The tubes were also a milestone on the way to the invention of Nixie lamps.

In the late 1920s and early 1930s, inventors realized that Geissler lamps could be modified slightly, using, for example, a suitably bent wire in the form of numbers or letters as a cathode. The author of one of the earliest patents for such a design was German inventors Hermann Pressler and Hans Richter. The Germans called it an “electric discharge lamp,” which acted as a “self-luminous sign. ” As the duo noted in their patent, granted in 1938, their lamp was more legible than a traditional neon sign and would have been cheaper to produce, since the glass tubes would not have to be bent so as to produce a specific sign.

However, on the other hand, they needed a glass casing to accommodate the gas, as well as the entire lamp structure, which was quite problematic and simply impractical. Not surprisingly, the invention did not find many takers. Today only a few pieces of such lamps have survived.

Patent US2142106 by H. P. Boswau (https://patentview.ip-tools.io/?numberlist=US2142106%2CUS2146576%2CUS2310328%2CUS2541041%2CUS2576099&mode=liveview)

However, another inventor had a better idea. Hans P. Boswau, a German immigrant who settled in the town of Lorain, Ohio, on Lake Erie, worked as chief engineer for the Lorain County Radio Corp. At some point, Boswau decided he needed a device to display numerical symbols and letters. On May 9, 1934, he filed two patent applications, which include the first full descriptions of what later became known as the Nixie lamp.

Boswau’s key innovation was to arrange the cathodes one behind the other in the same bulb, so that the cathodes could be switched individually. In this way, it was possible to build a single tube with 10 cathodes representing the digits 0 through 9. Interestingly, Boswau’s patents were issued to him, not his employer, so he probably did not intend to use them in his work.

Nowadays we are surrounded by all kinds of displays, it’s hard to imagine a time when this type practically didn’t exist. At the time when Boswau invented his lamp, most indicators were in the form of incandescent light bulbs and, in fact, the display of digits was not needed. It wasn’t until 1947, with the creation of the transistor and the birth of digital electronics, that the need arose to design a display capable of displaying digits.

Inditron GI-10 (https://www.industrialalchemy.org/articleview.php?item=423)

Several companies have stepped up to the plate. The first patent after Boswau for a numerical display was filed by Northrop Aircraft in June 1950. However, interestingly there is no evidence that the company created any lamp. The design was probably created only on paper. A more “real” candidate emerged in May 1954, when National Union Radio Corp. a well-known vacuum tube manufacturer unveiled its line of neon-filled indicator lamps, called the Inditron.

Like Boswau’s earlier invention, the Inditron consisted of a small glass bubble containing a stack of hand-curved digits that were electrically isolated from each other. The contacts of each digit were routed through a special seal on the outside of the tube. The digits served as a cathode, but there was no fixed anode. Instead, digits that were not in use acted as the anode. Let’s say the digit 7 was displayed. The negative terminal was connected to the egg lead, while the numbers 0, 1, 2, 3, 4, 5, 6, 8 and 9 had to be connected to the positive terminal. Needless to say, the circuitry required for continuous reconfiguration of the anode was very complicated.

Like other companies, Burroughs, a major manufacturer of calculators, computers and other business equipment based in Plymouth, Michigan, was also interested in numerical displays. However, it lacked the in-house expertise to develop them, so the company’s management decided to remedy that. In 1954, Burroughs acquired Haydu Brothers, a vacuum tube manufacturer in Plainfield, New Jersey, known for its high-precision manufacturing capabilities. It also hired engineer Saul Kuchinsky, a former National Union Radio employee who had worked on the Inditron, to run a laboratory at the Burroughs Research Center in Paoli . These two moves gave Burroughs everything it needed to develop the numerical indicator.

Haydu Brothers was founded in 1936 by two Hungarian-American brothers, George and Zoltan Haydu, and their father, John Haydu. It quickly gained a reputation for producing precision metal parts used in the construction of vacuum tubes. Burroughs wanted Haydu to produce high-precision lamps, in particular a line of beam-switching lamps bearing the trademark Trochotron. The beam-switching lamps were mainly used as high-speed counters, but were not capable of displaying numbers. For this purpose, Kuchinsky had to construct a new device. His team tackled a display lamp similar to the Inditron, but much improved. Meanwhile, Kuchinsky also served as general manager of the former Haydu plant in New Jersey, where he was able to oversee production.

One of the iconic Burroughs B7971 lamps (https://www.industrialalchemy.org/articleview.php?item=890)

Roger Wolfe, an engineer at Burroughs, recalls the first attempt to run the new lamp as follows, “We put the tube through a life test overnight. When we arrived the next day, so much cathode material had splashed on the dome of the tube that the digits were no longer visible. We invented a tube with a lifetime of 12 hours!”. After a few months, the team discovered that the addition of mercury vapor significantly extended the life of the tube. The sputtering was caused by accelerated neon ions hitting the cathode. But when the neon ions collided with the heavier mercury molecules, their energy dropped below the point where they could damage the cathode.

In August 1955, Burroughs unveiled its new indicator lamp at Wescon – the Western Electronic Show and Convention in California – which for many years was the leading electronics event in the US. Soon after, the company began shipping the first lamps to customers. The devices were mechanically superior to the numerical displays on offer at National Union: they had dedicated anodes made of wire mesh, and instead of hand-bent wires, the cathode digits were etched from thin sheet metal. The addition of mercury extended the life of the lamp, eventually to more than 200,000 hours. At the time, the lamps also gained their distinctive name. The first concept drawing was labeled “Numerical Indicator Experiment No. 1,” or NIX1 for short. In addition to being an engineer, Kuchinsky was also a good marketer and believed that successful brands often had a k or x in them (e.g. Kodak, Xerox, etc.). That’s why NIX1 was eventually renamed Nixie.

Quite quickly, there were many more companies producing Nixie tubes under license from Burroughs, which also supplied the tubes as OEM products that could be integrated into third-party systems. Nixie quickly became the go-to device for displaying numerical values in multimeters, meters and other scientific and industrial equipment. By the 1960s, more than two dozen companies in the United States, Europe, the Soviet Union, India, China and Japan were manufacturing and supplying Nixie lamps.

This period, from 1955 to 1960, was the golden age of the Nixie, but its successor was already in the pipeline. In 1962, Nick Holonyak Jr. working at General Electric Laboratories in Syracuse, New York, managed to tame the electroluminescent mixed phosphide and gallium arsenide crystal, and so the light-emitting diode was born, but it took several more years to turn the prototype into a marketable product.

How to control the Nixie lamp?

As I mentioned in today’s slightly longer introduction, Nixie tubes are currently experiencing a second youth. Nothing surprising about that, simply because they look quite nice, and many people, when building their own devices, choose them, abandoning contemporary designs.

However, most of the projects focus mainly on software issues, so I’d like to introduce you to two concepts for controlling Nixie lamps, but I’ll describe them more from the hardware side.

Transistor good for everything

Transistors that can control the Nixie lamp

The Nixie lamp itself is not a complicated device, having only a common anode and multiple cathodes corresponding to individual digits. Therefore, one of the simplest ways to control the lamp is to use a suitable transistor. Nixie can be plugged into the collector circuit and by controlling the base current accordingly, we are able to light the corresponding digits.

Schematic

The connection is quite simple – the tube is supplied with a high voltage (usually around 170V) through a current limiting resistor. The cathodes are connected to transistors, whose job is to close the circuit, when there is a voltage at the base. It may be puzzling why the entire tube has only one resistor at the anode, rather than a dedicated one for each cathode. The answer here is that we are unlikely to want to light multiple digits at the same time, in which case the lamp would be unreadable. Only a single digit will be lit at any given time, so the current flowing will always be the same and we do not need to use separate resistors for each cathode.

It is important to remember that in this type of design you can not use an “ordinary” transistor. Nixie tubes are powered by high voltage and the transistor must be adapted to it. The most important parameters here are the collector-emitter voltage and collector current. For example, the IN-12 lamp is powered by 170V, and its operating current is about 3mA. The transistor used in the project must be adjusted to these values. You can use, for example, BF458, BF459 or their Soviet counterpart KT604.

Dedicated Drivers

Soviet Nixie tube drivers

The second quite popular way to handle Nixie tubes is to use a dedicated driver. This is certainly a simplification of sorts, as you don’t need a stack of transistors, and you only need one specific chip.

There are several types of chips available on the market, the most popular are: 7441 and 74141. In addition, a Soviet design with the designation – KM155ID1 – is also quite common.

Schematic

The task of the circuit, as in the case of transistors, is to short-circuit the corresponding lamp cathode to ground, so that the corresponding mark will light up. Besides, the chip itself is a BCD-BIN decoder, so that four wires are enough to control a single lamp. The input of the chip is given a variable in the form of a BCD code, e.g. 1001, and then a low state will appear on the output Q9 of the chip, resulting in the illumination of the number nine.

Nixie's poisoned lamp

The so-called poisoned Nixie lamp

When buying Nixie lamps, you need to be aware that they are mostly no longer new. It may happen that the sign is not displayed correctly after startup. As in the photo above – the first vertical line of the letter H from the Hz sign is slightly darkened at the ends.

This is known as cathode poisoning and means that the lamp has worked for quite a long time before. But there is a way to salvage the lamp, you need to increase the current flowing through it until the sign is displayed correctly. You can increase the current by decreasing the value of the resistor connected to the anode. Once the lamp is shining correctly, you should leave it like this for several hours. Unfortunately, you should be aware that such action is not always effective.

Damaged Nixie lamp

Sometimes there are also lamps so depleted that nothing can be done anymore, but instead you can observe an interesting phenomenon that looks as if, the gas is trying to escape from the lamp.

Interesting Nixie lamps

Nixie lamps are not just parallel wires that allow the display of digits from 0 to 9. More complicated designs also appeared. One of them is a lamp designated MG-19B, somewhat resembling today’s 7-segment displays, except that it has an additional two segments so that several additional characters can be displayed.

MG-19B lamp (https://www.industrialalchemy.org/articleview.php?item=3129)
ITM2M (https://www.industrialalchemy.org/articleview.php?item=949)

One of the weirder Soviet inventions is the ITM2M lamp. It is a lamp based on the indirect gas discharge phenomenon and, like the Nixie, is filled with low-pressure ionizing gas, but that’s where all similarities end. The ITM2M lamps contain an array of 16 small thyratrons, each painted with one of four colored phosphors. Activation of one of the thyratrons causes ionization of the gas inside, resulting in illumination of the associated phosphor. These lamps have a rather complex power supply system, needing voltages of -300VDC, 150VDC and 75VDC to operate properly.

The Philips plant located in Eindhoven produced an interesting-looking lamp, designated Z550. In this design, the luminous segments were placed on a circle, against the lamp wall. Interestingly, this design was equipped with electronics that allowed us to select the digit we wanted to light with a 5V signal.

Philips Z550 (https://www.industrialalchemy.org/articleview.php?item=1116)

Also see:

Sources:

  • https://patentview.ip-tools.io/?numberlist=US2142106%2CUS2146576%2CUS2310328%2CUS2541041%2CUS2576099&mode=liveview
  • https://www.industrialalchemy.org/articleview.php?item=423
  • http://www.jb-electronics.de/html/elektronik/nixies/n_nixie_geschichte.htm?lang=en
  • https://www.hbomax.com/pl/pl/series/urn:hbo:series:GXJvkMAU0JIG6gAEAAAIo
  • https://www.industrialalchemy.org/articleview.php?item=890
  • https://www.industrialalchemy.org/articleview.php?item=3129
  • https://www.industrialalchemy.org/articleview.php?item=1116
  • https://www.industrialalchemy.org/articleview.php?item=949
  • https://spectrum.ieee.org/nixie-tube
  • https://en.wikipedia.org/wiki/Nixie_tube
  • https://www.steampunkalchemy.com/en/nixie-tubes

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top