Neon sign transformers differ from most other types of transformers one is likely to encounter. Unlike a more conventional transformer, for normal operation a neon sign transformer is specified to operate a minimum, as well as a maximum load. Why is this? To understand this requirement, we must look at what makes a neon sign transformer different from other types.
Neon tubes require a high voltage at a low current to operate. This power is supplied by a specialized transformer. Secondary voltages typically range from 1,000 to 15,000 volts, and secondary currents range from 20 to 60 milliamps (and higher, for large diameter “cold cathode” tubing). The current passing through a neon tube needs to be limited by some means, otherwise once the tube lights, the current will rise to an excessively high value. This regulation is accomplished by inserting a ferrous magnetic shunt (fig. 1) into the transformer core, such that the magnetic flux from the primary winding has an alternate (although high impedance) path around the secondary winding. As the current draw on the transformer secondary winding increases, more primary magnetic flux diverts1 through the magnetic shunt. While this gives the transformer a poor voltage regulation characteristic, it also tends to keep the neon tube operating current reasonably constant. We may electrically model this type of transformer as a conventional transformer which has an inductor in series with each of its high voltage secondary leads2 (fig. 2).
Note that some transformers may have more than one secondary winding and more than one secondary shunt. If a transformer has two secondary windings, the midpoint connection between the secondaries may be grounded to the transformer case. Depending on the exact configuration of shunts and secondary windings, a transformer may be referred to as having either a “balanced” (fig. 3) or an “unbalanced” (fig. 4) design, and this in turn determines what types of secondary wiring methods may be used. Refer to the transformer manufacturer’s literature for more details on this.
This constant current nature of neon sign transformers allows one to greatly vary the tube loading on a transformer. Unfortunately, one can radically misload a transformer and it will still appear to work, in the short term. Long term, transformer failure will usually result. For any given transformer, the tube load should fall within specified limits. It must not be too high or too low3. The question is, how does one determine the proper loading?
Methods of Determining Proper Loading
There are several answers to this question. The most commonly used method is by reference to manufacturer supplied loading charts. These charts indicate the minimum and maximum total length of neon tubing that may be used on a given transformer, as a function of the tube diameter, gas fill pressure and type of gas used (typically either straight neon or an argon/neon/mercury vapor mix). When using a chart, one deducts some amount of tube length for each pair of electrodes used (which occurs when multiple tubes are wired in series.) Other rules-of-thumb allow compensation for connecting tubes of varying diameters and gas fills in series. While this loading method seems straightforward and simple enough, and is certainly a good place to start in determining proper transformer loading, it does not always produce the desired results. This is because neon tubes may not exhibit the standard characteristics the loading charts are based on, due to processing variations and other factors. Therefore, we need to understand a little more about transformer characteristics, and look at some alternate loading techniques.
In the USA, neon sign transformers are rated primarily in terms of their open circuit secondary voltage and their short-circuit secondary current. While these ratings are specifically intended for use in calculating proper loading, they are also useful in determining if a neon sign transformer is functioning properly. The open circuit voltage may be measured with a good voltmeter, equipped with a high voltage probe. Typically, the higher voltage transformers (6000 volts and above) have their secondary winding midpoints grounded to the transformer case. The voltmeter common lead is connected to the case and a voltage measurement is made at either secondary terminal. The sum of these readings should equal the secondary rating. Be sure to measure the primary voltage, as the open circuit secondary voltage will vary proportionally with the primary voltage.
Secondary current may be measured by connecting an AC milliamp meter directly across the secondary terminals. The transformer will withstand this short circuit for a reasonable amount of time, because the magnetic shunts previously described allow what would otherwise be seen as abuse to be tolerated by the transformer primary winding. Again, the current reading should be reasonably close to the secondary short circuit value on the transformer rating plate.
Be sure to check the manufacturer’s literature for their specific recommendations on testing SGFP type transformers, as the above general test methods may not work properly.
This brings us to a second method of determining proper transformer loading. For US made transformers, the typical transformer secondary operating current is approximately 80 percent of the rated short circuit current. For a transformer rated at 30 mA, this would be approximately 24 mA. We can connect a milliamp meter in series with a transformer’s tube load and measure this current. Various manufacturers make high voltage milliamp meters specifically for this task.
I am told that in some areas of Europe, transformers are specified somewhat differently than in the USA. Typically, a transformer is rated in terms of its open circuit secondary voltage, its operating secondary current, and its “G” factor. The “G” factor is the desired ratio of loaded to open circuit secondary voltage. Typically, the “G” factor is around 0.5 (although this may vary), meaning for example that a transformer rated at 9000 volt open circuit should operate with a tube load connected at around 4500 volts. This may be measured conveniently with a voltmeter equipped with a high voltage probe. This is the preferred method of checking the loading on a European transformer, and in practice is somewhat easier and more sensitive than measuring the tube operating current. This method is starting to catch on in the USA as well. Unfortunately, US manufacturers don’t specify the “G” factor, but typically a value of 0.5 may be assumed.
One quick note here: when replacing a European transformer, the “G” factor must be taken into account. A replacement transformer with a different “G” factor will not operate a given tube load properly, even though its rated open circuit secondary voltage and operating current may be the same as the that of the original transformer.
Choosing a Loading Check Method
One would think that by using a loading chart in combination with measuring the secondary operating current and voltage that correct tube loading may always be determined. Unfortunately this is not always the case. There is one more variable in the mix: stray capacitance. This capacitance results from the close proximity of high voltage secondary wiring to its enclosing conduit (or other grounded metal objects), and between the neon tubing and the sign sheet metal. It is desirable to minimize this capacitance as much as possible. This means that any secondary wiring operating at high voltage relative to ground should be kept as short as practical, as capacitance varies directly with wire length. Various standardized wiring techniques, such as “mid-point return” and “virtual mid-point” have been devised to accomplish this goal. Both of these techniques involve keeping the wiring between the transformer and the electrodes on the first tube as short and direct as possible. Needless to say, the wiring techniques used must comply both with those methods allowed by the transformer manufacturer and with methods allowed by the NEC (for US installations), or with the appropriate European regulations for installations done there.
So what happens when capacitance intrudes in a neon installation? Two things, both bad. First, capacitance tends to counteract the current regulation of the neon sign transformer. Specifically, it tends to cancel the inductance in the secondary circuit of our previously described transformer model. This can cause excessive secondary, and thus, tube current. Unfortunately, installers tend to counteract this effect by adding additional tubing load to the transformer, or by selecting a smaller transformer. While this tends to restore the operating current, it is at the expense of operating voltage, which tends to rise excessively, leading to transformer failure. This is why both the operating voltage and current should be checked.
With some types of NEC 600-23(b) SGFP type transformers, the stray capacitance seen by either transformer secondary terminal should be similar. An imbalance may be interpreted by the protection circuitry inside the transformer as a fault, causing “nuisance tripping.” One of the best ways to avoid this situation is to insure that the GTO wires connected to either transformer secondary terminal be reasonably equal in length, as well as keeping them as short as possible.
Second, stray capacitance in combination with higher voltage transformers (typically units over 9000 volts) operating neon-filled tubes may invite transformer secondary circuit oscillations. These oscillations sometimes manifest themselves as flickering tubes and “buzzing” transformers, and are extremely destructive to both the transformer and to the high voltage wiring. Neither a voltmeter or a milliamp meter will conclusively detect this condition. Using an oscilloscope4 equipped with a high voltage probe, these oscillations may be easily seen. Other than reducing the capacitance as much as possible by mechanical methods, the best solution to this problem is to re-layout the sign wiring to use lower voltage transformers. This will, of course, require using more transformers to do the job.
Solid State Transformers
We have neglected to include so-called “solid state” transformers in our discussions of proper loading. This is because the operating frequencies used by these transformers preclude the use of regular voltmeters and milliamp meters, as they do not function well at these frequencies. The oscilloscope still works well, but few sign shops have them. On the other hand, most of these transformers are designed to operate with widely varying loads. When using these transformers, it is best to carefully follow the recommendations in the manufacturers’ data sheets.
Before publishing, this article has been passed for review to the members of the INA “Codes and Installation Forum.” I would like to thank everyone who reviewed this article for content and accuracy. Your assistance is greatly appreciated.
1 This is a bit of a simplification, but for purposes of illustration, it’s close enough.
2 Specifically, this is a model of a “balanced” midpoint grounded type of transformer. See figure 3 for a physical illustration.
3 A note on terminology: sign installers refer to a transformer as being “overloaded” or “underloaded” with tubing. Unfortunately, this refers to the length of tubing connected, not the electrical condition, which tends to be the opposite of that of the tubing. This is to say, when a transformer is “underloaded” with tubing, the transformer tends to draw excessive primary winding current.
4 Unfortunately, this type of equipment has been, in the past, both expensive and awkward to use in the field. Newer compact / portable equipment coming into the marketplace may make this measurement technique more practical in the future.