A panoramic survey of neon pumping systems
There are many different neon pumping systems available in today’s market. Here I evaluate the options, to facilitate understanding and choice.
My intention is not to give a scientifically precise description of manufacturing neon lamps or vacuum technique, but to help people who do not necessarily have technical and scientific backgrounds understand the basic principles involved when we leave the environment at atmospheric pressure and go into the peculiar space of vacuum, where the practical considerations that are valid in common empiric experience are no longer applicable. So please bear with the simplicity of my descriptions.
Neon pumping systems can be evaluated by paying attention to a number of factors: material used, the scheme of the plant, the quality and performance of the components, the ease of assembling, disassembling and maintenance, the possibility of automation, quality certifications and process control.
Glass or metal
Glass was the first material used in vacuum applications. Glass has low porosity and low outgassing. It is easy to clean, and highly resistant to chemical attacks (especially borosilicate, also called Pyrex). Glass is a good electrical insulator. Glass can also be easily and inexpensively shaped for all needs. In addition to this, glass is a familiar material for a neon glass bender. For all these reasons, neon manifolds (except for the mechanical pump) have been made of glass for many decades. The valves were made of bored double cones, ground and greased. Those valves were easy to break, especially when the internal vacuum pushed the cones against each other.
In Western Europe, glass was progressively abandoned for this reason, during the sixties, seventies and eighties, in favor of metals like copper, brass, aluminum, and steel, particularly stainless steel. Special rubbers with low outgassing properties have been created, to allow easy vacuum-tight connections between metallic parts. Brazilian rubber (for medium vacuums), Viton and Teflon for high and ultra vacuum applications, became standard in vacuum technology. International dimension standards have been introduced for flange connections so that pumps, instruments, gauges and conduits can be easily assembled.
In more recent years glass became convenient again. The new plastics, used in combination with metals, can also be very successfully used with glass. The glass valves now have Teflon plugs and Viton o-rings, and they never break during normal operations. Glass flanges perfectly meet the international standards for vacuum connection systems.
The only disadvantage of glass is its fragility, offset by a long list of advantages:
Glass is transparent and shows when its inner walls are dirty (heating and bombarding continuously generate vapors of impurities that condense on the colder manifold).
The porosity of the surface is lower so that it absorbs less and is easy to clean out.
Outgassing is less, and this reduces pumping time.
Glass, when cold, is electrically insulating, reducing the chances of electrical discharge through the manifold back toward the pump. Glass is safer in the presence of dangerous currents from the bombarding transformer (if perfectly grounded, metal is also safe). See Image 1.
Glass, especially borosilicate, is chemically inert.
Glass is a familiar material for the neon tube glassblower.
It is easier and cheaper for a neon shop to find leaks in the manifold. In metallic manifolds it is impossible to use the high frequency coils (a standard leak detector commonly used in every neon pumping system) for this and a very expensive Helium leak detector would be needed. This would force the neon maker to call and pay a specialized company.
Glass valves and connections are, in general, cheaper.
Image 1: The high voltage from the bombarding transformer finds an easy way (shorter and sometimes with lower pressure) toward the metallic grounded rotary pump – shown in turquoise. This path would be much shorter in the case of a metallic manifold. Where there is a long tube, the middle of the lamp, like at the ground, the voltage tends to 0 (yellow section). To avoid this it is advisable to connect the tubulated electrode to ground with a wire, or to bombard two lamps at the same time (if the transformer can supply enough voltage), so the two far electrodes will receive high voltage and the two tubulated electrodes connected to the manifold will have a voltage near 0 (if the lamps have similar lengths).
Plastics and rubbers
As explained above, plastics and rubbers make very easy and tight connections in a vacuum system. But, since many of those materials copiously outgas in very low-pressure conditions, a careful selection is needed.
Some materials, like Teflon, are found to have very low outgassing, more or less like glass. Teflon is also very resistant to chemical aggressions. For these reasons, and for its flexibility, Teflon tubing and components, instead of glass and stainless steel, are used in laboratory ultra vacuum plants. Modern glass valves also employ Teflon plugs.
O-rings for vacuum application are made of Viton, a special rubber designed for this use.
Brazilian rubber hoses are suitable until a pressure of 0.01 mbar (for example to connect the rotary vane pump).At lower pressures, like those created by a diffusion pump, this material starts to outgas and should not be used.
Materials like silicone copiously outgas and must not be used.
As a general rule we have to ensure that the material used is suitable for the vacuum created by the pumps. Using a pump with a high ultimate vacuum and introducing materials that outgas in such conditions is more damaging than having a pump that provides a poor vacuum. A relatively poor pump will leave impurities that will react in the lamps, creating an inert material (if such impurities are not too much); material outgassed from rubber, plastics, or oils and greases may create long lasting wiggling, bad stains, or contaminate the electrode activation.
To conclude, this is the principle that we must remember: the higher the ultimate vacuum, the more careful we need to be in selecting the suitable materials.
Oils and greases
As for plastics and rubbers, we must pay attention to the characteristics of oils and greases when they are under vacuum. For the rotary vane pump, the oil must have the suitable viscosity, indicated by the manufacturer of the pump, and a vapor tension (a function between pressure and temperature) lower than the ultimate vacuum of the pump at its operating temperature. That means, in practice, that the oil must not generate vapors in the vacuum conditions created by the pump.
The old glass valves and stopcocks required frequent applications of silicone grease to lubricate the double cones. The new valves and connections can provide a perfectly vacuum-tight connection without being lubricated, but a very light application of silicone grease to the o-ring will protect them.
Diffusion pumps, in the neon field, use silicone oil as a propeller. At low pressure and at the suitable temperature, the oil creates high-speed jets of vapors that are immediately condensed and recovered in the heating vessel. The oils can differ in density, viscosity, boiling temperature, and ultimate vacuum. The oils that have a lower boiling temperature start to work at a higher pressure and reach a lower ultimate vacuum.
The design of the neon manifold
The neon manifold will be our companion for a long time and must not only provide good neon but also be able to keep up with possible future habit changes, technology, and market demands. For these reasons it should be able to adequately process neon tubes of very different dimensions, fill them with different rare gases and mixtures, allow for easy maintenance, and allow the introduction of automatic controls.
Ultimate vacuum and pumping speed
It is important to have a general understanding of neon tube processing and its physical principles.
Neon tubes are heated by an internal electrical discharge in air at low pressure (2-8 mbar). The combination of heat, radiation, ionic bombardment and low pressure aids in the release of impurities trapped in the first layers of the inner glass walls or fluorescent powders and allows us to pump them out. Without this cleaning, the lamps would not be able to operate for more than a few minutes.
In addition to this, the bombarding process accomplishes another task: the activation of the electrodes. The inside surface of the shells is coated by a layer of chemical compounds that must react, leaving oxides of Barium, Strontium or Calcium. Only after this chemical reaction can the electrodes work at the appropriate current.
In order to accomplish this double job, we need means or gauges to monitor the pressure and the current in the lamps. If the pressure or the current are not appropriate, the lamps can be seriously damaged and will have a much shorter life. A well-made neon lamp should be able to work for ten years or more of continuous operation, but this primarily depends on the conditions of the bombarding process.
Since this process frees the impurities by the combined effect of a lot of energy in the form of heat conducted, heat radiated and a number of other radiations created by the electric discharge in air, and by the low pressure, the pumping system must pull them out as fast as possible and as much as possible, before the lamps cool down.
The impurities that remain in the tubes, once they are cooled down to ambient temperature, will be partially condensed on the inner walls and reabsorbed again. So, the vacuum system has a limited time for pumping out. This means that only a certain degree of vacuum can be reached during this interval, and after that the lamp must be immediately filled with the rare gas and tipped off. This also means that it is useless to have equipment that gives an ultimate pressure much lower than this level. Let’s see why, looking at Image 2 below:
When a gas is compressed, like in the atmosphere, it tends to instantly occupy all the empty volume, and, when the pump works, it continuously creates empty spaces. In the situation on the left, the atoms will all go toward the opening of the narrow conduit that represents the small glass tube connecting the lamp to the manifold, the tubulation. The single atom can only move in the direction where all surrounding atoms go. It will be impossible for it to move against the stream created by the movement of the mass of air heading toward the empty vessels. In a few seconds, 99% of the air will be transferred to the pump. But then the situation changes going closer and closer to the situation on the right. At very low pressures, like the ones needed to produce a clean neon tube, the atoms are moving freely in all directions and they find the narrow opening by chance, and by chance move through the tubulation to the pump. So, in those conditions, the speed of the pumping does not depend on the ultimate vacuum, but mainly on the geometry of the system. A wide straight tubulation would lead to a faster pumping, but we cannot increase the diameter of the tubulation because it will become extremely difficult or impossible to tip off the lamp from the manifold once the process is completed. Once melted by the flame, a wider glass tube will collapse under the action of the external air pressure, and then it will crack because of the accumulated stresses.
Image 2: On the left we see the scheme of a neon tube connected to the manifold before pumping.
On the right, see the situation at the end of pumping. The air atoms are represented by circles.
In fact, the speed of the high vacuum pumps is rated according to the section of their inlet openings. So, in the final stages of our pumping, the section and the length of our tubulation will determine the speed at which the vacuum is produced in our lamps, no matter how fast our high vacuum pump is working. In practice, the efficiency of our system will be able to increase the pumping speed until the pressure is more or less 0.05 mbar (5 x 10-2 mbar). If our set of pumps can produce a higher vacuum, only with time will this vacuum be created inside our lamps. The higher the vacuum we want, the longer, exponentially, will be the time that we will have to wait.
In practice, a vacuum in the lamps of 0.001 mbar is a realistic target. Our vacuum gauge, normally located in the manifold, will not be able to read inside the lamps and will indicate a lower pressure than that.
The above discussion helps us to understand the great help that Helium can give in cleaning neon tubes.
Helium is one of the five nobles gases, so it is chemically inert. Its atom is the smallest after Hydrogen. If, after bombarding, once the pressure in the lamp goes down to 0.1 mbar, we let some Helium into the lamp, we can mix this gas with the remaining impurities and pump them all out together. The remaining gas at the end will be composed mainly of Helium, a rare gas that will not have any negative effect. In other words, Helium can be used as a rinsing gas.
Helium has other interesting characteristics:
It has a high ionization potential, which makes an electric discharge through this gas very hot, and the surrounding glass easily heated.
The cathode hit by the light positive ions of Helium remains colder during an eventual second bombardment, while the electrodes, which have already been activated, do not need to be heated again.
A well-designed neon pumping system should provide a way to add this rare gas, in order to profit from its positive qualities.
Activating the electrodes
The chemical reaction that burns the original internal coating of the electrode shell takes place at a high temperature (900°C) and under the effect of the ionic bombardment of the cathode. The residue of this burning is a mix of oxides of Barium, Strontium and Calcium, which is the effective active surface of the electrodes. This thin layer of oxide is essential for the operation of the activated electrodes. We need to be able to accurately control the pressure during the final stages of the bombardment, when the shells must reach the needed temperature. At this critical moment a strong electrical discharge with too little pressure can cause evaporation of the activation mass. The electrode would then operate as a non-activated one and it would not withstand the current for which it was designed. The life of the lamp would be, in this case, very short.
The appropriate pumps
The goal of the pumps connected to the manifold is to provide the right ultimate vacuum and the fastest speed in a particular interval of pressures, between 1 mbar (the end of bombarding) and something lower than 0.001, which is the pressure that we can reach in the lamp in its short cooling down time. As we have seen before, an ultimate pressure lower than 0.00001 mbar (1 x 10-5 mbar) is not our target and does not provide any benefit. On the other hand, an ultimate pressure that cannot go lower than 0.001 mbar, if it can provide acceptable neon lamps in most jobs, can be critical in certain situations (long or big lamps, small diameter glass like 6 or 7 mm). Pumping speed is greatly reduced when approaching the pump’s ultimate pressure, and this would happen right in the interval where we need the fastest pumping. In addition, the general performance of mechanical pumps is reduced when the pumps and their oil get old. For this reasons a well-designed system must include a secondary high vacuum pump.
The primary pump
A double stage rotary vane pump can provide an ultimate vacuum of approximately 0.005 mbar. This pump’s speed depends on its chambers’ sizes. Increasing the size of this pump in order to avoid the installation of a secondary high vacuum pump is an option. However, this would make it more difficult to control the pumping speed at higher pressures. Unless a fine valve is introduced in order to control the air flux from the lamp, this would happen at around 1 mbar, when the electrodes are burning their activating material at the end of bombardment. But, since this would reduce the pumping speed of the system, it would be a contradictory design.
From the above considerations it becomes clear that the best design should physically include two tools, one for the bombarding pressure, with a fine control of the pumping speed, and one for the high vacuum to be reached after bombardment, while the tube cools down and we need the highest possible conductance.
A double stage rotary vane pump with a pumping speed of about 5 m3/h (at atmospheric pressure) seems to be ideal as a primary pump. This machine would be able to pump 4 3-meter-long 25 mm lamps down to bombarding pressure in about 15 seconds.
The modern rotary vane pumps include an anti-suck back system or valve. Without this, the oil of the pump would be pushed by atmospheric pressure into the evacuated manifold when the pump is turned off. If an old pump is used, an automatically activated valve should be installed at the inlet opening of the pump in order to let air inside the tube connecting the manifold.
The secondary high vacuum pump
For the high vacuum pump we have to be conscious of the fact that, if the ultimate vacuum can vary from one type of pump to another, the speed at the lowest pressures, as we have seen before, mainly depends on the diameter of the inlet opening. Since we have to pump through narrow tubulations, all the available high vacuum pumps are generally oversized for our needs.
The choice could be among three types of machines:
A silicone oil diffusion pump
This family of pumps uses special oils. The oils themselves act as the pumps’ engines. Mercury has been almost completely abandoned because of its toxicity. The oil is heated to a temperature between 160° and 180°C (depending on the type of oil chosen), where it boils if subjected to a pressure around 0.05 mbar. The boiling temperature depends on the type of oil chosen.
The graphic in Image 3 below shows that, at the pressure where we need the fastest operation, the pumping speed of an oil diffusion pump is rather low (see the gap between the descending blue curve of the primary pump and the rising red curve). For this reason it is important that the pump is maintained and installed in ideal conditions to fully develop its function.
In general oils that boil at higher temperatures can reach a higher ultimate vacuum, but also require lower pressure to start working, and are not preferred in our application.
Silicone oils are the ones that better withstand oxidization and among them we find the types that start sooner, at higher pressure, and are more suitable for our purposes.
Because of their relatively low cost, and very easy maintenance, steel diffusion pumps are used in the neon industry. In scientific laboratories, on the other hand, diffusion pumps of small dimensions have largely been replaced by molecular pumps.
Both metallic and glass diffusion pumps are available. The vacuum industry offers a range of three-stage diffusion vacuum pumps, which produce a higher ultimate vacuum than the single stage represented in Image 3, but this characteristic is useless for us. Their pumping speed remains the same and, remember, our lamps almost exclusively depend on the diameter and size of the electrode tubulation. Optimally, metallic diffusion pumps require cooling by water or forced ventilated cool air.
The single stage glass diffusion pumps have a simpler design and more space available for cooling down. They perfectly condense the oil (which reduces possible oil back-streaming) and do not require water cooling or forced ventilation. An important aspect of oil diffusion pumps is the control of the energy given to the oil to boil. As in any boiling material, the temperature depends on the pressure, and the oil will be protected from overheating if it constantly remains under vacuum. So, if the heater gives out too much energy, this will produce excess vapors. Those vapors will condense too late and the pump will not work efficiently. Electronic thermo regulators that receive information from a thermocouple fitted in the oil vessel of a glass diffusion pump can ensure a perfect equilibrium and the best efficiency. This will also adjust the energy given to the heater on the variation of the ambient temperature.
A molecular pump
These pumps have propellers, or rotors, turning at a very high speed and controlled by quite sophisticated electronic circuits. Among them we can distinguish three families: turbo-molecular pumps, hybrid turbo-molecular drag pumps and molecular drag pumps.
Image 3 shows that turbo-molecular pumps are the latest to start pumping and, consequently, they are not preferred in a neon pumping system.Hybrid turbo-molecular drag pumps and drag pumps, especially the latter, are very suitable. Compared to diffusion pumps they start to operate sooner, at higher pressure, and reduce the possibility of contamination from oil back-streaming.
Because of their sophisticated mechanics and their very high rotating speed, molecular pumps can be instantly destroyed if a small solid grain enters through the pump’s rotors. A suitable special filter must be installed to protect the machine. You need periodic and regular maintenance, and lubrication of the bearing. Because of their relatively lower rotation speed, drag pumps are less problematic from this point of view.
Image 3: Comparison of the speed of different types of pumps related to their maximum capacity and to the pressure in the vacuum system.
Connection of the secondary high vacuum pump
In cases where a molecular high vacuum pump is installed, the manifold must be designed in a way that avoids the inrush of air at atmospheric pressure to the molecular pump. This would seriously damage the high speed rotating mechanics. Especially with turbo molecular pumps, you need an automatic system to prevent the opening of the inlet and outlet valves if the pressure in the manifold is not lower than a certain limit. A drag pump, which rotates at lower speed and has stronger mechanics, can forgive quite a few operator mistakes.
The diffusion pump with oil as its engine does not have this problem. But the continuous passage of air at atmospheric pressure will end up oxidizing the oil and absorbing more dirty vapors. This would reduce the pumping speed, and so it requires more frequent maintenance and oil replacement. The scheme of the manifold below can, maybe, reduce the cost of the system but does not get the best performance from the diffusion pump. All the air and impurities in the lamps will go out through the diffusion pump. Two separate paths for the bombarding pressure and for the high vacuum are not provided. The operator could maintain the valve S just slightly open while processing the lamps and would maneuver the valve B. He or she would get great help to control the pressure and will avoid a too-low bombarding pressure that would damage the electrodes. To get the high vacuum they would then completely open both S and B.
The two alternative ways to the lamps will allow the operator to connect the next set of lamps while the previously bombarded ones are cooling down, if there is space in the neon shop.
Image 4: Neon vacuum system with single pumping way
The following scheme represents an improvement, since it provides a separate valve and an additional primary pump for the high vacuum diffusion pump, which will remain protected from atmospheric pressure. Here it would also be possible to install a molecular pump instead of the diffusion pump, while in the scheme above (Image 4), the molecular pump would be rapidly destroyed by the continuous air inrush.
The conduit with B, or the valve itself, should be reduced in section, or have a fine control of the flux of air in order to better control the bombarding pressure. Unfortunately, most commercial manifolds do not have this characteristic and install big expensive valves in this position. Smaller valves would save costs and do a better job.
These first two schemes are generally offered by American companies.
Image 5: Neon vacuum system with double pumping ways (two rotary vane pumps)
In Western Europe the traditional scheme used is the following (Image 6):
Image 6: Neon vacuum system with bypass valve
This scheme, in a very simple way, allows us to avoid the installation of the second rotary vane pump exclusively dedicated to produce vacuum for the secondary high vacuum pump. Having installed the valve F, we can connect the manifold to a sole primary pump in two alternative ways:
Through the high vacuum pump, opening H and F;
Through the valve B, bypassing the high vacuum pump, which will be maintained constantly under vacuum having closed H and F.
Here, the same points made earlier about the size or characteristics of the valve B or its conduit are applicable.
The following scheme (Image 7) represents an improvement to the traditional West-European and is my favorite:
Image 7: Neon vacuum system with bypass valve, vacuum reservoir and separation valve
As you can see, the improvements consist of:
- An additional gas supply path dedicated to Helium;
- A separation valve S, (coming from the lamps) located after the entrance of the conduit from the bypass B (while in the American scheme it is positioned before);
- A vacuum reservoir positioned at the exhaust opening of the diffusion pump.
The function of the separation valve S
This scheme’s function is to maintain almost the whole manifold constantly under vacuum. When connecting the lamps, S will be closed, together with all the other valves except A. Then, A will be closed and B opened. The lamps will be evacuated until more or less 1 mbar (the changing sound of the rotary pump will tell the operator that there is enough vacuum in the tubes). Only at that point will S be opened and the pressure will be monitored by the vacuum gauges during the bombarding process and the following high vacuum pumping. Not exposing the whole manifold to atmospheric pressure, we get two advantages: 1) we avoid absorption of air with its impurities by the inner surfaces of the system, reducing the needed pumping time; 2) we avoid damaging the head of the vacuum gauges. The head of the Pirani gauge, for example, has an incandescent metallic spiral that, if oxidized because of exposure to atmospheric air, would not allow a correct reading.
The function of the vacuum reservoir
The vacuum reservoir’s job is to constantly provide the diffusion pump with enough vacuum to operate at higher speeds. After bombarding, in a few seconds the pressure should have dropped to about 0.1 mbar. Because of the large volume of the vacuum reservoir, opening H and F will cause the pressure to drop almost instantly to about 0.03 mbar, near the pressure of the highest speed of the diffusion pump, similar to the gap between the blue and red line in Image 3. If a drag or hybrid turbo-drag molecular pump is installed, the presence of the vacuum reservoir is not important.
Image 8: Glass vacuum reservoir
Rare gas transfer systems
Ways of supplying rare gases — this is a hot topic, and there are many options of gases and mixtures. The blue gas mixture can contain different percentages of Neon-Argon, and parts of Krypton or Helium, in order to adapt the lamps better to different climatic conditions, or for indoor or outdoor installations; Krypton or Xenon, or mixtures containing those gases are also used for dimming effects, special colors, or to avoid use of Mercury.
Different solutions are available for gas bottles and for controlling the gas flux to the manifold and the lamps:
The traditional glass bottles
These, containing 1.25 or 2.5 liters of gas at atmospheric pressure, are completely obsolete. They present big disadvantages: the cost of the gas itself becomes very expensive because of the amount of labor needed to produce the glass bottle, to pump out the air and to clean, fill and seal it. And all this only for 1.25 liters! It is also impossible to use a bottle without losing all its contents. These bottles require the installation in the manifold of a couple of glass valves or a single needle-metering valve.
The metallic heavy bottles under high pressure (from 50 to 150 bar)
These are the most economic for mass production. They require a significant initial investment for the high-pressure bottle itself and for the pressure reducer that must be vacuum-tight. This system does not allow an easy, fast and cheap interchangeability of different gases or mixtures. Transportation and stocking of this type of bottle requires special care because of the potential danger from the high pressures. This system also requires the installation of needle valves to control the flux of gas to the manifold.
The metallic light bottles under medium pressure (maximum 12 bar)
Compared to glass bottles the cost of the gas is greatly reduced (about 4 times). This solution also offers the best and easiest interchangeability: a bottle can be removed and replaced in a minute, without losing rare gas. No significant investment is needed to purchase the bottles and connect them to the manifold. A simple and cheap metering needle valve is used for a double function: a tight connection to the bottle, and to open it, accurately controlling the gas flux when the lamps must be filled with rare gas. No other valve or pressure reducer is required.
Image 9: Rare gases bottles
Three parameters must be continuously checked during bombardment:
Different types of electronic gauges are available, with different degrees of precision. The catalogs of the main companies selling vacuum equipment have many choices. What is important is that the gauge must measure the total absolute pressure, with the same reading for all types of gas. In fact this gauge will be also used for the filling pressure of the rare gas.
Mechanical capsule gauges are very practical and reliable, but not very precise (tolerance of about +/- 10%). All these gauges, especially the electronic ones, must be regularly controlled and maintained. A wrong reading in the filling pressure will produce enormous damage to the reliability of the lamps. A neon shop with a wrong reading of the filling pressure can produce bad lamps for 6 months before the first lamps die, indicating that something is wrong. But at this point, the market will have already received a quantity of products that will not only damage the reputation of the neon shop, but also the image of neon lamps in general. A wrongly filled neon lamp will last weeks or months instead of decades!
More traditional ways to control bombarding and filling pressure are Mercury gauges or U oil gauges. The great advantage of these types of gauges is that they cannot be wrong! The disadvantages are, for the U gauge, that a valve has to be opened before letting atmospheric pressure go into the manifold, and closed once a good vacuum is reached. (In the scheme in Image 7, the separation valve S, which avoids atmospheric pressure in the manifold, makes this unnecessary).
In the automatic pumping systems, these traditional gauges cannot be used because they do not send electric or digital messages. But, even in this case, they should be used as instruments for easy and frequent comparative control of the electric gauges.
The current in the lamps
Checking the current flowing into the lamps during the bombarding process is very important, because the heat transferred to the glass walls and to the electrodes depends not only on the pressure and quality of the gases developed, but also on the current intensity. A standard milliamperometer can be connected in series to the lamps, but it must stay at a safe distance from the operator. If a manifold includes a milliamperometer in the main rack, a specific electric circuit must be dedicated in order to reduce the voltage from the secondary coil of the bombarding transformer to a safe low voltage.
The temperature of the lamps
The lamps must reach, during bombardment, a temperature of at least 220°C. To check this parameter a thermocouple thermometer is commonly used. This also must stay at a safe distance from the operator. The negative point of this gauge is the fact that the clamp between the thermocouple and the lamp will absorb a certain amount of heat while it is measuring the temperature. The effect is that the reading is something lower than the actual temperature. The amount of this difference depends on the dimensions of the clamps and on the material used. Infrared thermometers are also very suitable and can measure without touching the glass.
I’ve seen many operators processing lamps without using any way to check the glass temperature. This is a great mistake, considering that a simple piece of paper gives precious indications about the temperature: paper starts to get brown at 180°C, gets black at around 220°C and burns at about 300°C. It is reliable and cheap — why not use it?
A typical Pirani gauge cannot indicate the filling pressure, since its reading depends on the nature of the gas measured. So its function is to give indications on the tightness of the vacuum system, including the lamps, and on the efficiency of the set of pumps. The reading scale goes down to 0.001 or 0.0001 mbars.
A Voltmeter measuring the secondary voltage of the processed lamp can give indications on the quality of gases or vapors that are developed. If, for example, traces of Mercury are present, this will reduce the voltage needed for a certain length, diameter and pressure of glass tubing.
The use of modern vacuum connections allows the creation of systems that are made of assembled parts. A properly designed neon manifold allows an easy and rapid disassembly. This is very important, because frequent cleaning of the pumping system is crucial to the quality of the lamps.
Two types of connections are used in neon manifolds:
A Viton o-ring is located between a male threaded tube and a female threaded cap. Screwing the cap onto the tube will compress the o-ring onto the smaller tube introduced. When this smaller tube has a diameter bigger than 7 mm, means must be designed to prevent external pressure from pushing this tube into the bigger one.
A centered o-ring is compressed between two flanges. With this type of connection, the diameter of the joined tubes can remain the same. A manifold assembled with flanges is stronger and more precise. It is also easier to disassemble and to put together again. Flange connections also follow international standards and are more suitable for connecting gauges and pumps.
Compression fittingsGlass and metal flangesGlass flanges
Image 10: Glass and metal connections
In the scheme above we see valves with different functions:
The valves marked B must regulate bombarding pressure and should provide a smooth and gradual opening to prevent undesirable pulling out of the mercury and the coating phosphors. As described above, in the first two schemes this target can be achieved with the combined action of the valves B and S. Restrictions on the conduit can also be used to get to this target.
The valve A (air inlet) also must provide a smooth and gradual opening.
The valves H and F should be quickly opened and closed. They also should not create restrictions on the high vacuum conduit, to maintain maximum pumping.
The material of the valves can be metal or glass (mainly borosilicate).
The bombarding transformer
This transformer must provide enough voltage and a strong current in a length of tubing containing few mbar of air. 18 or 20 KV at 700 mA are enough to process a length of two lamps of 3 meters each in 25 mm diameter tubing.
The transformer should be finely calibrated in order to adjust the power to a wide range of lengths and diameters. Many options are available, from magnetic chokes, big variacs, electronic dimming circuits. The last ones are the most suitable for automatic control systems.
The high voltage and current provided by the bombarding transformer is extremely dangerous. Every metallic part must be grounded. Means must be installed to prevent accidental contact with live parts. This can be achieved with physical barriers or with barriers of photo cells, with automatic disconnection of the transformer voltage when they are crossed.
You must take care to avoid the risk of conductive objects in the area of the tubes and the high voltage wires, and lead out the high voltage.
Mercury is dangerous and toxic. It remains permanently embedded in living organisms. A well-designed manifold must have efficient mercury traps to prevent dispersion of this toxic material into the manifold. The traps should be easy to clean.
Image 13: Safety barriers
Automating your vacuum system can help you make sure your lamps turn out correctly despite human errors. It is also useful when you need quality certification.
Opening and closing valves, filling rare gas, controlling the bombarding current -all these tasks can be automated. The bombarding transformer and both metallic and glass valves can easily be automated. The automation of both bombardment and vacuum procedures is new.
Always remember, however, that neon manufacturing depends on many varying continuous conditions. Automation can never replace the most important factor: an operator with deep technical knowledge and good instincts