2. Nixies presentation and clock construction
Before dwelling into the technical specifications, here is a short introduction to nixies, as well as some material on the clock case construction itself. No real link between the two subjects, but they were both too short to have their own page.
Page summary2.1 What are nixie tubes ? 2.1.1 Theory of operation 2.1.2 IN-18 specifications 2.1.3 Anode resistance 2.2 Clock case construction
2.1 What are nixie tubes ?
Nixie tubes are a type of gas-filled tube which are used to display symbols (our our case, digits). The anode of such a tube composed of a mesh that covers most of the tube's inner area ; the cathodes have the shape of a digit. Reminder: the anode is the positive electrode (where electrons arrive) ; the cathode is the negative electrode (where electrons leave). The tube is filled with a very low-pressure gas mixture composed mainly of neon. When a certain voltage is applied between the anode and one of the cathodes, a visible glow appears around the selected cathode, i.e. the selected digit is lit.
While nixies were convenient in the valve era, when most electronics was made at high voltages, this technology is now outdated, replaced mostly by LED displays or LCD screens. Not only nixies require high-voltage driving circuits, but they also have quite a list of failure modes. The most common ones include the simple breaking of the glass envelope (releases the special atmosphere inside the tube, making it unusable), and cathode poisoning (which causes dark zones on the digits over time). Cathode poisoning occurs when some digits remain turned off for a long time while other glow. The sputtered material (see below) from the lit digits will form a resistive coating on the unused digits. This coating dims the affected zone and can mask entire sections of the digits in more severe cases.
Despite all this, nixies retain some appeal because of their originality and beauty.
Notice the subtle purple tinge on the sides of the tubes ; this is caused by the small fraction of mercury present in the gas mixture, supposedly added to reduce cathode poisoning.
2.1.1 Theory of operation
Nixie tubes belong to the cold-cathode tube family, as opposed to hot-cathode tubes. In the latter type of devices, the electrons are expelled from the cathode via thermionic emission, i.e. the cathode is brought to a high temperature so that electrons are emitted due to thermal agitation. On the other hand, cold-cathode devices rely simply on the voltage applied between the anode and the cathode (which is usually higher than in hot-cathode tubes) to extract the electrons. The resulting current is induced either directly by the electrical field or by other electrons coming from previous collisions of the initial one (secondary emissions). While cold-cathode devices are indeed much colder (about 40°C at most) than their hot counterpart (several hundreds of °C), some heat is always produced, for example simply by the current crossing them.
But how do nixies actually work ? Long story short : de-excitation of sputtered atoms. The relatively high voltage between the anode and the cathode (around 180 V) ionises some of the gas atoms present in the tube. These positively charged ions will accelerate towards the negatively charge cathode. Some of these ions will be able to get close to the cathode where they will collide with (neutral) gas atoms, which will in turn hit the cathode. If their kinetic energy is sufficient, they will be able to rip off a metal atoms from the cathode (this is sputtering). At this point, electrons snatched from the cathode will collide with these metal atoms and will excite them. And, as they de-excite, it is a well-known fact that they'll emit a photon for energy conservation. These photons constitute the glow we see.
I've eluded some aspects to keep things short. A more complete explanation of nixie operating theory can be found on the corresponding ExplainThatStuff page.
Figure 2.3 — Close-up on a lit digit of a nixie. One can clearly see the characteristic glow of these tubes (which has, by the way, absolutely nothing to do with incandescence).
Some precisions on the last part. It is a consequence of quantum mechanics that the angular momentum of, in our case, an electron in an atom is quantified, i.e. only a discrete set of energies is allowed to him (as opposed to a continuous set). So, what I referred to as the excitation of an atom is actually the placement of one of its electrons to a higher energy level. After staying a short time on this state (which is inversely proportional to the energy difference between the states), the electron will fall back to its original level and will return the excess energy by emitting a photon. The frequency ν of this photon is given by ν = ΔE/h, where ΔE is the difference in energy between those two level and h is Planck's constant. The exact explanation on this mechanism (an electron - photon interaction) is a matter of quantum electrodynamics (QED), which is beyond the scope of this page.
2.1.2 IN-18 specifications
The nixie tube I chose to use on my clock is the awesome IN-18, the largest model produced in the USSR, with very fine and elegant 40mm high digits. IN-18 are still available on the Web for about 30€ each. They can be supplied with a 170 to 180V tension and a 4 to 4.5mA current for optimal conditions. More informations (and nice pictures) can be found on Tube-Tester.com. The complete datasheet can be downloaded in PDF format from the link from here.
2.1.3 Anode resistor
When wiring a nixie tube to a power supply, it is imperative to use an anode resistor to limit the current going through the tube. This current will determine the digits brightness and too much current will greatly reduce the nixies life. The required resistance is usually a few kΩ ; I used 10kΩ carbon resistors for this clock. An online nixie anode resistance calculator can be found there.
Figure 2.4 — Complete nixie wiring schematic with resistance calculation example.
[Image credit: M. Moorrees]
2.2 Clock case construction
All pieces were entirely cut by hand. The case is made entirely in wood and metal (aluminium and steel). The main section, in darker varnish, contains the two Arduino's and the breadboard for all the connections (tubes and inputs). The lateral casings with oblique tops contain the speakers, which have been salvaged from an old cathodic television (yes they're big, but I pretty much need that to wake up properly).
The front decorations, i.e. the eagle and the rays, are made of 0.5mm thick aluminium sheet. The rays were additionally sanded to give them a matte finish. The top panel, which supports the nixies, is composed of two wooden plates assembled with glue and is fixed to the main section with magnets. An aluminium cover with brushed aspect has been placed on the nixies' bed. Rubber layers have been glued on the lower side of clock main support plate (the one with steel reinforcing borders) for better stability. Last but not least, the SD card holder as well as the aluminium ornamental plate are located on the backside (see this picture). A folded metal bar has been added below the card holder to hide the wires going into the case.
Here are a few pictures during the construction.
Let's talk about the nixie supports. I first decided I would made them myself, and I tried to drill the adequate pattern (six copies) on the wooden plate itself, as it can be seen on fig. 2.5 (second picture). Then I put power chord-type copper wire in these hole for connection. I assumed this would, at one time, offer a good connection and a good stability by forcing the nixies inside a little bit... This was a very bad idea, as the drilled patterns weren't exactly matching the nixies ones. The result was that either nixies couldn't be correctly aligned with all of its connection, either they broke as a result of trying to force them in an inadequate pattern. After breaking two nixies like this, I removed these handmade supports placed purpose-made Bakelite supports instead, which worked just fine (see fig. 2.6).
Figure 2.6 — The backside of the top plate. On this picture, the wires connecting the Bakelite supports to the breadboard are currently being soldered. On the background, one can see the inputs with their connecting wires as well. On the foreground is the fixation magnet.
On next page, we'll have a look at the wiring diagrams of the clock as well as the electronic schematics of the boards used.