This page is under construction. As time permits, articles and neon shop tips that were published previously will be placed here. Technical articles or shop safety or helpful hints are always welcome. The information in these articles is believed to be reliable, but has not been independently verified. Readers should make their own checks or tests and check for updates and changes before relying on the information presented.


[email protected]


January, 2000 Association Newsletter

February, 2000 Association Newsletter

March, 2000 Association Newsletter

April, 2000 Association Newsletter


Excerpts from the January, 2000 Association Newsletter

By:    Bobby Howard

Safety Reports

  An economical cleanup powder for accidental mercury spills is now being offered direct from the manufacture, Action Technologies. The product is named HgX®, a water-soluble powder that can be used to decontaminate neon shop floors, walls, bench tops, and other surfaces.

  The product can be mixed with water and sprayed or wiped, or applied in the dry form. It can also be mixed with saturated sawdust for sweeping floors to control dust and mercury vapor hazards. This reasonably priced product is available from:

Action Technologies,
100 Thompson St.,
Pittson, PA 18640

  Another product that is reported to work very well is “HgAsorb”. It is available from Lab Safety Supply, who also offers complete mercury spill cleanup kits. Mercury Spill Control Station, #8B-20754,  would serve a larger neon shop very well. A smaller version is their Mercury Cleanup System, #8B-20876. Both of these kits also include Hg Indicator Powder, which changes color in the presence of mercury. Both kits are  reasonably priced as well. Contact Lab Safety Supply, 800-356-0783, Fax 800-543-9910, or check their website at

  Proper disposal of contaminated materials should be of high concern, so check with your local authorities for a hazardous material recycler in your area. INA recommends that all shops have the proper mercury cleanup kits readily available and training in their use.

  - - - - - - - -

Neon Bending Cloth

 TecnoLux, Inc., a manufacturer of glass tubing and equipment for the neon and cold cathode industry, announces a new Neon Bending Cloth . This new cloth is unique in the neon industry and is ideally suited for neon patterns, tabletop covering and bending jigs. TecnoLux says it offers significant advantages over other “non-asbestos” fabrics and it is virtually smoke, discoloration, and itch free.

  Using textile grade fiberglass, the material is woven and impregnated with PTFE (Teflon®). This creates a fabric that is strong, easily cleaned, and chemical resistant. The fabric possesses high tensile strength, low friction, and good resistance to cut-through. The PTFE coating on the fabric creates a non-stick, flame resistant surface that also stands up to aging and degradation. The Neon Bending Cloth is available in 54" X 10-yard ( item # 427-54-10 ) and 38" X 20 yard rolls ( item # 427-38-20 ).

  Along with the comprehensive range of TecnoLux neon and cold cathode tubing and accessories, the new Neon Bending Cloth is available through sign supply distributors worldwide. You can also contact

 TecnoLux, Inc.
103 14th. Street,
Brooklyn, New York, 11215, USA
 Telephone (718) 369-2845
 Email : [email protected]

  We’ve been using this new cloth for a number of months now, and find it’s very resistant to burning, and you won’t have that itchy skin problem anymore. In doing repetitive letters on this cloth, the wear factor is excellent.

- - - - - - - -
- - - - - - - -

Vacuum Measurement Units
and Scientific Notation

By Telford Dorr

 When we measure very low pressures, as found inside neon tubes during the pumping process, we speak of these pressures in terms of absolute pressure. This is the pressure of a gas as compared to a perfect vacuum (which would have a numeric pressure value of '0'). This is different from the more common pressure measurement method used to measure the pressure in compressed air tanks. This pressure is referred to as 'gauge' pressure, and is measured relative to atmospheric pressure.  In a sense, there is no such thing as a vacuum. What we commonly refer to as a vacuum is simply an area of pressure below that of the normal atmosphere.

 As the first experiments involving vacuum were conducted with mercury barometers, one of the unit of pressure measurement used was 'inches of mercury'. It was discovered that our normal atmospheric pressure would support a column of mercury in a barometer approximately 30 inches tall or approximately 760 mm. In neon work in the USA, we measure the low pressure inside tubes in the unit of 'torr (1 mm Hg).  This system of units works well for measuring filling pressures of our tubes, as the range of normal fill pressures is from about 4 to 18 torr. However, it doesn't work as well for measuring the pressures needed to insure a clean and empty tube before the rare gas is inserted into it. For this, we use the unit of the micron, where 1000 microns = 1 Torr.

 In areas of Europe, the standard unit of pressure used is the 'bar', where atmospheric pressure is around 1.013 bar. This unit is a little large for vacuum work, so a more commonly used unit is the millibar, where atmospheric pressure is around 1013 millibars. As you can see, the unit 'millibar' is close in value to the 'torr'. You can convert between them by using the following conversion factors:

1 torr = 1.333 millibar or 1 millibar = 0.750 torr

 When we get involved with vacuum systems using secondary pumps, such as diffusion pumps, we find even the 'micron' can be too large. Rather than invent yet another unit of pressure measurement, we revert to what is called 'scientific notation', and express all pressures in terms of 'torr'. The value of this technique is that we can easily express an extremely wide range of measurement. This is because this notation involves using exponents, which in this case is a multiplier factor expressed as a 'power of ten'. For example, the number 1000 can be expressed as '10 to the third power', which means '10 times 10 times 10'. This can be written as 10exponent 3 or 10e3. Example:

1234.5 = 1.2345 x 10exponent 3 = 'one point two three times ten to the third power'

 We can also express numbers using negative exponents. Example:

0.00123 = 1.23 x 10exponent -3 = 1.23 x 10e-3

 Typically, the main number (or mantissa) is expressed with one significant digit to the left of the decimal point and everything else to the right. A trick to remember: to convert a number to scientific notation, move the decimal point either right or left such that it is to the right of the first significant (e.g. non-zero) digit. The number of positions you move it is the numeric value of the exponent. If you move it to the left, the exponent is positive; to the right, it is negative. Try this on the examples above. Simple!

 How do we apply this knowledge to neon work? Suppose we have a pumping system which has a diffusion pump, and suppose we have a cold cathode type discharge gauge attached to our manifold, as well as the more common thermocouple gauge. My discharge gauge has a two scales, both calibrated in 'torr'. The 'high' scale has a range of 10e -5 to 10e -3. What does this mean? Well, per our discussion above, we see that 10e-5 torr is the same as 0.00001 torr, and 10e -3 torr is the same as 0.001 torr. Therefore, my discharge gauge reads from 0.00001 to 0.001 torr. As my thermocouple gauge indicates down to 1 millitorr (or 'micron'), which is the same as 0.001 torr, we see that the discharge gauge picks up nicely there the thermocouple gauge quits.

 Now suppose I want to convert a measurement from one system of units to another. I can do this by multiplying the number by the proper conversion factor. For example, suppose my discharge gauge indicates a pressure of 1.23 x 104 torr, and I want to convert this to millibars. Scientific notation makes this easy. Example:

1.23 x 104 torr x 1.333 millibars per torr = 1.64 x 104 millibars

 You create the mantissa value (the main number) by multiplying the two numbers together: 1.23 x 1.333 = 1.64. You create the exponent by adding the two exponent values together: 4 + 0 = 4. Note that the conversion factor can be expressed as 1.333 x 100, which is where the '0' came from.

 By knowing the principles of scientific notation, we can compare the scale calibrations of different gauges to each other, using the same basic unit of measurement. We can also easily convert from one system of units to another by multiplying by the proper conversion factor.

Return to Index

Excerpts from the February, 2000 Association Newsletter

Reports From Europe

By Cor Zorn

 The biggest difference between the American and the European method of making  neon is probably the coating techniques. In America, pre-coated tubes are almost exclusively used, but in Holland for example, it's just the opposite. Most bending is done with clear glass, and then powder coated after bending

 There are two common methods, dry coating and wet coating. With dry coating, there are two methods of applying the binder, with a sponge, or with glass beads. Both methods start by thoroughly cleaning and etching the inner surface of the glass with very aggressive chemicals, followed by a water flush. The remaining water is removed with alcohol or acetone and  then dried with a hot air stream.  The units are now ready for the binder.

 Using the sponge method, a piece of sponge cut like a cylinder roughly 1 1/2 the diameter of the glass is wetted with the binder and squeezed dry between the fingers. This part is important because too wet is just as bad as too dry. It's all a matter of experience. Using compressed, oil free air, the sponge is forced to travel the inside of the unit back and forth a few times, leaving a thin layer of binder.

 The second technique uses glass beads to apply the binder inside the tube. A bottle filled with glass beads of approx. 2 mm diameter is wetted, drop by drop with the binder to get just the right amount. The prepared beads are then put into the unit to be coated. There should be enough space left to shake them around. This also leaves a very thin layer of binder inside the tube.

 It’s now time to  apply the dry coating powder. The powder is ladled into the tube usually by a adapted teaspoon or the like. An inch of powder (measured in the tube) is sufficient. By shaking and turning the unit, the powder is allowed to coat all inner surface of the tube. A little tapping with a wooden stick  also helps. After a second pass through the tube, excess powder is removed and the unit is baked, electrodes applied, and then pumped as usual.

  This all seems like a lot of work but an experienced person can do a lot of powdering in an hour. It depends on the complexity of the units but an average of 10 to 15 units is possible.  And just think, you only have to stock inexpensive clear glass to make all common colors.

 Next time we’ll look at wet coating techniques.

- - - - - - - -


 Marking your neon units with a number, the date, or your trade mark, isn't that hard. Durability of that mark is the hard part. I've used all kinds of markers and found one which gives a nice looking and very durable mark. I have used a metallic marker NIJI (NJN 93 made in Japan) with great success, but I think other  brands may perform equally well. They are available in two colors, silver and gold. Once applied, the mark stays on no matter what.

 To accomplish that, the mark needs to be put on before bombarding. The heat from bombarding makes the bond between the glass and the metallic part of the ink very strong. When testing the ink, I literally melted a piece of glass in the flame without removing the mark.

 Your units will have a very professional appearance by using a stamp wetted with the metallic ink from the marker and quickly pressed on the unit. The label on the marker warns about using on soft polyvinyl chloride plastics so make sure the material of the stamp is capable of withstanding the solvents in the ink.

- - - - - - - -
- - - - - - - -

Vacuum Measurement  (continued)

By Telford Dorr

 This month, let’s look at various types of medium pressure electronic vacuum gauges, and how they work. This information is important to know, as various types of gauges are suitable for use in neon work, and others are not, depending on where and how they’re used.

 It should be readily apparent that some type of vacuum measuring device can assist one in producing more consistent and higher quality tubes. While I recognize that there are those out there who can produce a fine tube without any gauges whatsoever (and I applaud such skill), I feel that gauges give one the certainty that they are doing so, as well as indicating in advance when things are not well with the manifold (as happens from time to time...)  On the other hand,  having gauges on your manifold will not compensate for a lack of pumping experience - you still need to understand what the gauges are telling you.

 The most common electronic gauges found on neon manifolds are the thermocouple gauge and the Pirani gauge. The basis of operation of both gauges is similar, and depend on the thermal conductivity of gasses at low pressure.  Basically, as the pressure of a gas is reduced, so is its ability to dissipate heat. In both gauge tubes, a heated filament is surrounded by the manifold gas. The temperature of this filament is determined by the input power to the filament and the rate of heat loss from the filament to the surrounding gas.

 In the thermocouple gauge, there is a thermocouple welded to the center of the filament. By measuring the voltage produced by the thermocouple, we know the temperature of the filament.  If we keep the input power to the filament constant, then the filament temperature is determined by its heat loss to the surrounding gas - the more gas, the greater the heat loss. Therefore, we can calibrate what amounts to a filament temperature meter to directly indicate pressure.

 In the Pirani gauge, there is no thermocouple. Instead, we exploit another characteristic of a heated filament - that its resistance varies with temperature.  If we arrange the circuitry connected to the filament properly (generally in the form of a ‘Wheatstone bridge’), we can simultaneously heat the filament and measure its resistance.  Again, this resistance meter is calibrated to indicate pressure directly.

 There are a few caveats to thermal pressure measurement. [1] It only works if the composition of the gas is known, as thermal conductivity varies radically with the type of gas surrounding the filament.  Generally, these gauges are calibrated for a nitrogen atmosphere (air is mostly nitrogen.)  As such, one cannot use one of these gauges to measure the pressure of the neon or argon backfill gas (unless the gauge has special scales for doing so.) Their main value is in indicating how effectively the roughing pump is doing its job. [2] The range of pressure measurement is limited. Generally, thermocouple gauges are good down to about 1 micron, and a bit lower for the Pirani. An additional note: generally the accuracy of these gauges is limited, and can only be expected to be accurate within a factor of two.

 There are a few considerations to be made when choosing and installing one of these gauges.  It must be recognized that the neon shop manifold can be a hostile place for a gauge tube to exist. The main hazard to be dealt with is flashback from the bombarder.  A bombarder arc can fry both the gauge tube and/or any associated electronics.  There are two methods of dealing with this problem.  The first is to use a battery operated gauge.  This generally eliminates any conductive path from the gauge back to ground, as there is no connection to ground via a power cord.  However, there could be one through the manifold operator if he was to touch the gauge when the bombarder is active.  As such, one should be extremely cautious, as bombarders generally take no prisoners.

 The second method of protecting the gauge is to precede the gauge tube with a grounded electrode (we’re speaking of a glass manifold here - if you use a metal manifold, you should already have it securely grounded.)  Of course, you can/should use both methods, for extra security.  My manifold has a grounding electrode preceding the gauge tube (and the stopcocks), and I ground the gauge tube shell as well.  Of course, if your bombarder has a propensity for flashback, you may have to come up with alternate solutions, such as isolating the gauge tube from the manifold with a stopcock.  Various manufacturers make gauges with protected circuitry for use in a neon manifold environment.

 I have heard that some gauge tubes contain gold in their internal construction. These types should probably be avoided, as mercury combines with gold, and any mercury contamination within the manifold could ‘poison’ the gauge tube, destroying its accuracy.

 For those of you interested in such things, there are articles on making your own gauge readout indicators (and for the truly adventurous, making your own Pirani tubes) in the publication The Bell Jar.  While I’m not advocating making your own equipment here, as good equipment is readily available on the surplus market at reasonable prices, it’s interesting to experiment with this stuff to gain a further understanding of how this equipment works.

  [1] pg 82, Building Scientific Apparatus, by Moore, Davis, and Coplan, published by Addison Wesley.  ISBN 0-201-13189-7. This is a good general reference on constructing vacuum equipment.

  [2] This publication devotes itself to all types of experimental vacuum physics. For more information on the publication The Bell Jar, see on the Internet.

Return to Index

Excerpts from the March, 2000 Association Newsletter

Powder Coating (continued)

By Cor Zorn

 In the previous article I promised to discuss the wet powder technique. This method, which is also called suspension coating—or the Phillips technique, was developed by the crew of Phillips, Eindhoven in the early fifties. The principle is quite simple, mix powder with binder and solvent, pour it in the tube, let the excess run out, dry the tube, bake it out, add the electrodes, and pump the unit. Easier said than done you’ll say if you ever get the chance to actually try it.

 First of all, the tube needs to be clean and dry just like for dry coating, and the tube must be the same temperature as the powder mix. The operator must wear wool gloves to prevent heat transfer from the fingers. Mix from the bottle is poured into the unit and time becomes a factor now! The time needed to powder one unit must be roughly the same for every unit. The unit is then picked up and the mix is allowed to cover the entire inner surface. When that's done the excess is poured out. Neither too much or too little, just enough to form a thick strip at the front of the unit which is still with the ends pointing up. When the operator is satisfied with the distribution of the mix collected at the front, the unit is turned face up (ends pointing down) allowing the powder mix to cover the surface with a second layer.  If there is too much mix in the unit, a “too heavy” strip will build. Too little will cause streaks and  partial absence of the second layer.

 The next step is to dry the unit with a hot air blower. Carefully move one of the ends of the unit well above the hot air stream, then lower it slowly as the powder starts to dry.  After a few minutes the powder layer has enough strength to remain on the blower until it is completely dry. There should be an very good exhaust system in this room to remove any vapors.

 After the final visual check, units are baked in an oven at 450 *C (840 *F). They should be positioned ends down so the air can pass through the units as much as possible. This causes the binder to evaporate completely, leaving an extremely clean tube. Basically the wet coating technique leaves nothing in the tube but powder. Usually it's much cleaner than with the dry coating technique where the binder may not have evaporated completely. A good exhaust system on the oven is also important. Also, the 450 *C temperature is near the strain relief point of soda glass. The oven should cool down gradually, and the units are now ready for electroding and pumping.

 With this method, the only equipment needed is a motor driven roller bank where the bottles of suspension roll constantly in order to maintain a good mixture. The ready-made suspensions are of course, more expensive, but very convenient to use. Preparing your own suspension is also possible but it requires careful milling of the powder to the exact particle size. The powder used for making liquid suspensions is much finer than for dry coating, but if milled too fine the suspension has a reduced light output.

All data needs to be recorded so the next batch has similar properties. An experienced operator can easily wet-coat 30 or more units per hour depending on the complexity of the units. The diameter of glass suitable for wet coating is the same as for any other technique. One liter of suspension can coat up to 250 meter of glass. This can vary depending on the diameter and the skill of the operator.

Due to the extreme cleanness and the well balanced mix, the light output and longevity of a wet coated unit will usually exceed all other methods of  making neon units. There are, however, also drawbacks to this process. It's much more time consuming than using pre-coated glass, and the back of the unit usually has a thick layer which makes it only suitable where the back isn't seen.

 Often we hear that it would be more expensive to coat your own, but that's hardly true. After the initial investment in an oven and perhaps a roller mill, the savings can be substantial.  If you compare pre-coated  with post-coated prices, materials only, the ratio will be about 1.6 to 1. In other words it's much cheaper,  especially when you have a high volume shop and have a person occupied with making and applying suspension. And from an environmental point of view, the only waste you have is clear glass.

- - - - - - - -
- - - - - - - -

Electronic High Vacuum Gauges

By Telford Dorr

 This month, we’ll investigate high vacuum gauges. These gauges pick up where the thermocouple and Pirani gauges discussed last month leave off - at pressures below 10e-3 torr (1 micron). If you have made the investment in a diffusion pump for your manifold, you need some type of high vacuum gauge to know if it’s working or not, and if so, how fast.

 The classical method of determining high vacuum in a neon manifold is to build a high vacuum indicator tube. You weld a pair of electrodes onto a short (8 inches) stick of glass (regular green works well).  Tubulate it in its center, and connect the tubulation to the manifold. Typically, this indicator tube is excited with a 2 - 3 kv, 20 ma core and coil type transformer. In use, as the pressure in the manifold drops, the tube starts to light up. As the pressure drops further, the light output of the indicator tube reaches a peak, then decreases and eventually goes out altogether. If you can’t get the indicator tube to extinguish, it’s a pretty good bet that the vacuum in the manifold is poor.  The advantage of this method is that (1) it’s rugged and (2) you can fabricate it yourself. The problem with this method is that you don’t know what the pressure is when the light goes out, nor do you know how much lower the pressure goes afterwards. To solve this problem requires investment in a high vacuum gauge.

 There are several types of gauge sensors capable of measuring pressure at low levels. Unfortunately, most are not suitable for neon use, for one reason or another. The classical sensor for laboratory high vacuum work is the Bayard-Alpert thermo ionic ionization gauge (BAG)1. This is available as either a glass tube, which is connected to a port on the vacuum system, or a nude version of the sensor, which is basically just the inner guts of the tube, which is inserted directly into a vacuum chamber. The advantage of the second type is that it may be rebuilt when the filament burns out. On the other hand, it is very susceptible to mechanical damage. These sensors work well on (and are designed for) systems which are maintained continuously at high vacuum. Unfortunately, an inrush of air will burn out the filament. Also, a minimum vacuum level is required before the tube can even be powered up. Because of this, these sensors are not suitable for neon work.

 The type of sensor which will work on a neon manifold is the Penning cold cathode ionization gauge2. These sensors have a small insulated chamber which contains a magnetically confined discharge. The sensor tube has a (approx. 15 mm) metal tube extending from its body which is connected to the manifold with an o-ring type compression fitting. The tube is excited by a high DC voltage (2 to 4 kv, current limited by a high value resistor). The current passing through the tube is measured and displayed on a meter calibrated in pressure. As the pressure drops, the current flow through the tube decreases. My unit (an old surplus Televac) has two scales (10e-7 to 10e-5 torr, and 10e-5 to 10e-3 torr). The sensitivity is selected by a front panel switch. The advantage of the Penning tube is (1) there is no filament to burn out, and (2) they can tolerate exposure to atmospheric pressure (at least for a while - although I don’t turn mine on until I open the valve and put the diff pump online). As an additional advantage, I have found several of these types of units available in surplus shops.

1 See pg 82, Building Scientific Apparatus, by Moore, Davis, and Coplan, published by Addison-Wesley, 1989

2  See pg 159, Neon Techniques and Handling, by Samuel Miller, published by Signs of the Times Pub. Co, 1977, for a somewhat dated example of one of these gauges.

Return to Index

Excerpts from the April, 2000 Association Newsletter

Processing Neon Tubes

By Cor Zorn

 Basically there are two methods used to process a neon tube - the bombarding method and the oven method. Actually there is a third method, using a bushy flame, but that isn't used a lot. There are not many neon companies who actually use the oven pumping method, and the ones who do are mostly located in Holland.

 I learned the oven technique at Super Neon in Holland in the early seventies. Essentially it's a way to pump multiple units in one batch.  There isn't a bombarder involved. The electrodes are converted by high frequency current and the tube is heated in an oven.

 Picture a large rectangular box lined with fire resistant insulation material. Electric heating elements are placed in the lid or at the bottom of the oven. Some distance from the bottom, there is a grid of heavy walled pyrex tubes forming a second "floor". This is where the units are placed when pumping. Along the sides of the oven are holes big enough to put a mercury trap through. These holes are tilted a little, going up from the outside in, to prevent hot air from escaping and heating and evaporating the mercury outside in the traps. All around the oven is a glass tube with tubulations for every hole in the oven wall - a mega-manifold, so to speak. This manifold is divided into sections with big valves to facilitate leak hunting and for connecting to the pumping station. Actually, this kind of oven also doubles as a baking oven for the wet powdering process, but is then used at a higher temperature.

 The next piece of equipment needed is the high frequency (HF) electrode heater. This is in fact a transmitter, like a radio broadcast transmitter. The output isn't going to an aerial but into a small water cooled coil of six or seven turns which can be slipped over the electrode.  The one I used at Super Neon was a Phillips model capable of providing 1000 watts of HF energy - enough to heat a electrode in less then 6 seconds. In fact, you have to pass the dumet electrode wires rather quickly to prevent heating them and causing a crack at the electrode's head.

 The pumping station is similar to the ones used for bombarding - same setup but with bigger pumps.  Because there is relatively more pumping time compared to the bombarding process, they aren't that much bigger. I remember the setup at Super Neon used a two stage mechanical pump and a Hg 45 mercury diffusion pump, capable of 45 liter/sec. at 10-3 Torr. That was quite sufficient for the usual 35 to 50 units present in one batch.

 The actual processing starts off with the neon units getting a extension of the tubulation, first bent down along the body of the electrode and  then bent again well below the shell. This double bend is necessary to be able to put the HF heating coil over the electrode later on.  Another connection method used is to put a side tubulation below the electrode, like v/d Ven still does.

 The prepared units are now placed in the oven and the mercury trap is sealed on (both to the unit and to the outside manifold). When done, the roughing pump is allowed to evacuate the units while the oven is heated to about 150 *C (300 *F).  This evaporates the majority of the moisture present in the tubes.

 After this initial heating, the oven lid is removed and the diffusion pump is put to work. Now, it’s time to convert the electrodes one by one, starting at the far end of each unit.  When done, the lid is put back on and the oven is heated to about 400 *C (750 *F).  This temperature is chosen because, unlike the baking out of wet powder, now there is a force of one kilo per square centimeter present (14 pounds / square inch). When the glass is heated beyond the relief temperature, the presence of that force causes permanent stress.

 When done heating, the oven is gradually opened, allowing it to cool down.  The units are filled with noble gas and tipped off at the manifold, leaving the mercury trap on. The mercury is dumped in and the unit is tipped off at the electrode. When there are different filling pressures needed, because of short and long units, the short ones are filled first and tipped off. The excess gas is pumped away and the rest can be tipped off.

 The whole process, one batch, takes about 4 hours to complete. The person operating the oven spends 3 hours of that time actually working.

One of the major advantages of oven pumping is that it is a uniform and easy repeatable process. This is also one of the major disadvantages - when there is something wrong, all units of that batch are affected.  Luckily, that rarely happens unnoticed and irreversible by a good operator.

 There are some really good pictures of the oven process on Kenny Greenberg’s web site - for those with access to the internet. Also I have found an old photo of myself at the oven which is available to see at:

- - - - - - - -
- - - - - - - -

Spurious Operation of Open-circuit Protective Devices in Circuits Operating Neon-filled Tubes Powered by HF Converters

By Don Dunthorne*

 Since the introduction of transformers with integral earth leakage (‘ground fault’) and open-circuit protective devices, we have noticed a large increase in the incidence of spurious tripping, particularly of the open-circuit protection.  The problem mainly occurs in circuits with pure neon gas filled tubes and is associated with damp mornings. It occurs in transformers of different   manufacturers. The problem is fairly widespread, and we must assume that it also occurs, beside  Europe, in other parts of world.

 A few observations can be made.

1. Because of the high turnover voltage - the point where the discharge characteristic in the tube changes from a positive to a negative resistance (or the moment current starts flowing) - in neon  tubes, the re-ignition of the tube on each half cycle can be uncertain.  It needs only a relatively small reduction of the supply voltage to the converter (‘solid-state’ transformer) for the tube to flicker, or fail to restart on some half cycles.

2. Under normal circumstances, the self-capacitance of the transformer secondary windings stores energy which is released during the ‘off’ period between half cycles, when the unit is electrically off. This energy causes a ringing oscillation to occur, and the oscillatory voltage is added to the normal waveform and ensures a reliable restart each half cycle. This is the reason why there is rarely much reported evidence of instability in neon circuits.

3. However, it only needs a very small leakage of current between the high-voltage terminals or  between the terminals and earth to dampen this oscillation. Even a leakage current of 2.5 mA,  much too low to normally activate the earth-leakage device, will cause the oscillatory voltage to disappear.

4. If this happens, say, on a damp morning, then a previously stable neon circuit may become unstable and flicker. Flicker can often mean that the tubes extinguish for several cycles of the main supply. The ‘off’ period can easily reach 200 msec, causing the open-circuit protection device to trip.

5. Before the introduction of open-circuit protection, an unstable flickering circuit would quickly have dried out and therefore stabilized.  However spurious tripping of a protective device results in expense for the sign maker and will cause problems with the customer.

 All this can be a real problem. Sign companies are experiencing difficulties with customers who are complaining of signs which are often not alight.

   There is real benefit in having open-circuit protection in small signs, like window signs, and in neon units installed within arms reach. Switching off a sign if a tube is broken is a good safety feature. Also on indoor signs, or residential neon, there is rarely a problem with leakage due to moisture. It is on outdoor signs where this problem occurs. Signs installed above arms reach (which, in practice, nearly always means outdoor signs) are the most vulnerable to spurious tripping.


 Installing HF converters (‘solid-state’ transformers) with open-circuit protection for pure neon filled tubing outdoors can cause serious trouble. Also the cable length and the enclosure the converter is build in can cause troubles.  Last but not least, the higher the output voltage, the higher the chance there is for spurious tripping of the device.

* Mr. Dunthorne is a member of Cenelec and an  officer in the British Sign Association.

Return to Index