An Experiment to Investigate Stresses on

Neon Sign Transformers in

Tesla Coil Primary Circuits

Terry Fritz

2-28-98

Summary:

It is not unusual for neon sign transformers to fail when used in Tesla coil primary circuits. There are two popular methods of connecting the transformer to the primary circuit. The first is to connect the transformer across the spark gap. The second method is to connect the transformer across the primary capacitor. It is generally believed that connecting the transformer across the spark gap places less stress on the transformer. This paper will show the method used to determine that this belief is well founded. This experiment shows that if the transformer is connected across the primary capacitor, it will see much higher voltage and current stresses than if it is placed across the spark gap. The nature of these stresses is shown which provides valuable insights when attempting to design transformer protection circuits.

Description of Experiment:

The experiment is divided into two parts. In the first part the primary circuit is tested with the transformer connected across the spark gap. Then, in the second part, the primary circuit is tested with the transformer connected across the capacitor.

There is no secondary system in place for this test. Data suggests that not having the secondary system in place causes greater stresses on the primary components. Also, having the extremely high voltages and associated arcing of a secondary circuit would endanger the test equipment used in this experiment. It is felt that the results were not materially affected by not having a secondary system in place.

In order to protect the test transformer, the test voltages were reduced to approximately 2500 volts RMS at the transformer. This lower voltage allows for testing under all conditions while presenting minimal risk of transformer damage. The results found at this lower voltage are believed to be very similar to conditions at higher voltages. Only the magnitudes are believed to change.

Experimental Setup:

Figure 1 shows the connection of the various components used in the experiment. It can be seen that the primary circuit is of typical design. Fiber-optically coupled voltage and current sensors are connected to a digital-storage oscilloscope and a laptop computer that is used to store the waveforms.

Neon Transformer

The neon sign transformer is a Transco Inc. catalog number S1512. It is rated for 15,000 Volt, 60mA output with a 120 VAC primary. It is new and in excellent condition.

Spark Gap

The spark gap is a simple stationary electrode type. Two 1/2 inch diameter steel rods with polished ends are mounted to give a gap of 0.02 inches. The center of each electrode has a small hole through which air is pumped from a small aquarium style air pump. This provides cooling of the electrodes and helps the quenching of the gap. The gap is shown in Figure 2.

Capacitor

The primary capacitor consists of ten 1.7nF 30,000 Volt ceramic "doorknob" style capacitors mounted in parallel between two brass plates. The capacitor assembly is shown in Figure 3.

Inductor

The primary inductor is made from 1/4 inch copper tubing. It is a linear coil, 12 inches in diameter and 12 inches long. There are 24 turns spaced 1/2 inch apart. The value of the inductor is 120uH. The inductor is shown in figure 4.

Variac and AC control Equipment

The transformer is driven by a typical 1KVA variac. There are also fuses, line filters, meters, etc. which are used to monitor and control the power applied to the transformer. This system is shown in Figure 5 for reference.

Oscilloscope and Computer

The oscilloscope is a Tektronix TDS210. It is connected to a Toshiba laptop computer which is used to download graphic files to disk. This setup is shown in Figure 6.

Figure 1.

Figure 2. Spark gap. Note air tubes to electrodes.

Figure 3. Primary capacitor assembly.

Figure 4. Primary inductor.

Figure 5. AC control equipment.

Figure 6. Oscilloscope and computer.

Fiber-optic voltage and current measurement

The voltages and currents associated with the transformer are measured with fiber-optically coupled voltage and current probes. The voltage probe consists of a string of 50, 200K ohm surface mount resistors which provide a current to the transmitter that is proportional to the voltage being measured. The current to the transformer is measured by a 10 ohm resistor array that is connected to another transmitter. These signals are sent to a receiver that converts the light signals back to a voltage that is measured by the oscilloscope. The receiver can be seen to the left of the oscilloscope in Figure 6. The voltage divider array and current shunt are shown in figures 7 and 8. Figure 9 shows the complete primary setup.

Figure 7. The voltage divider array.

Figure 8. The current shunt and transmitter.

Figure 9. The complete primary system.

Results:

In the first test, the transformer was connected across the spark gap as is commonly recommended. The voltage and current waveforms for this situation are shown below.

Top trace Voltage 2000 volts/division.

Bottom trace Current 50mA/division.

The first picture shows the voltage (top) and current (bottom) waveforms. The voltage waveforms shows the 60Hz output of the transformer with the gap firing and dropping the voltage back to zero. This produces a single fast 90 mA current spike at each discharge as is shown in the expanded view of a single discharge in the second picture.

In the second test, The transformer is connected across the primary capacitor. Similar oscilloscope pictures are shown below in this situation.

Top trace Voltage 2000 volts/division.

Bottom trace Current 50mA/division.

The stresses in the first picture look somewhat similar to the first experiment. However, When the waveform is expanded it can be seen that the transformer is experiencing heavy current oscillations. There is also a significant voltage oscillation that is also being forced onto the transformers output terminals.

Conclusion:

The experiment shows that when the transformer is placed across the primary capacitor, it will experience very heavy current oscillations and significant voltage oscillations. When the transformer is place across the spark gap, the voltage simply drops to zero and there is a single current spike, which also simply drops to zero. A transformer is placed at great risk when it is connected across the primary capacitor. This supports the common assumptions and observations.