Restoration of the Griffith Observatory Tesla Coil

By
R.G. Aurandt, et al

Since 1936, visitors to the Griffith Observatory in Los Angeles have been impressed by the conical Tesla coil located in the facilities Hall of Science. Up through the mid 1960's, the coil worked quite well by providing a spark four feet in length. In the years that followed, the resonator's performance had declined to a short brush charge. Even in this condition, it would demonstrate an electric field radiation by lighting up unconnected gas filled lamps. However, the snake-like blue-white spark was sadly absent.

This decline had been noted by many including four TCBA members on the west coast, namely Ed Angell, John Foster, Victor Hymowitz, and myself. We shared ideas as to what might be done about the situation and hit upon the possibility of volunteering our services for repairing the equipment. A meeting with Wayne Nelson, the Observatory's Technical Director, aided in getting our ideas communicated to Dr. E.C. Krupp, the Director of the Observatory. Dr. Krupp consented and the project was both started and completed in the month of June, 1985. It seems most appropriate that the coil be brought back to its former operating condition as this marked the Observatory's 50th anniversary. All of the work was done on Saturday mornings before the Observatory was opened to the public.

Perhaps this is a good time to bring out a bit of history of the Observatory's Tesla coil. The 1/2-wave unit on display is actually one-half of the two-coil 1/2-wave system constructed in the 1920's by Mr. Earle L. Ovington. In those days, the coil was used for stage (vaudeville) demonstrations. Later, Mr. Ovington gave the coil to his friend and associate, Dr. Frederick Finch Strong. In 1936, Dr. Strong donated the coil to the Observatory. The two-coil oscillator and the bank of Leyden jars can be seen in Photo #1.

Only 6 of the 12 jars can be seen because the wooden rack was two rows deep. The jars were 0.01 mfd. each for a total of 0.12 mfd. Nothing is known regarding the transformer used in the photo. But the low frequency transformer currently used is the one that came with the coils. It is not of the current limiting type. The data plate states, “Made by Roentgen Transformers, 2000 watts, secondary volts 20,000, primary volts 55, 110, 220.” In an attempt to limit the current draw, the four 55-volt independent primaries had been hooked up for 220 volts but had been supplied with the normal 1930's line voltage of 110 volts. The transformer is presently fed 117 volts with a measured output of 11,000. Because of its physical size (15"X30"X16", oil filled, weighing 250 pounds), the unit could easily supply 10 KW for short periods. Don't forget, in the good old days, electrical equipment ratings were grossly reduced so that units would not wear out. Progress? hmmm....

The two-coil system when delivered to the Observatory was without primary coils and corona caps. The coils were first tested in the basement of the Observatory using temporary primary coils and all twelve jars. The system produced an intense spark of over eight feet in length. The initial tests were conducted under the direction of Mr. Leon Hall with the assistance of Mr. Kenneth Strickfaden. Readers may remember strickfaden's artistry in the high voltage scenes of the Frankenstein series and other science fiction movies done at Universal studios in the 1930's.

Lack of space prevented public demonstration of both units so only one coil was put on display. This set-up employed five of the twelve Leyden jars. This provided a spectacular display for short periods on a 20-ampere circuit, but during the many years of service, the Leyden jar bank suffered individual breakdown. Although they were replaced with the spare units, the supply eventually ran out and the resonator had to be kept alive by making adjustments in the primary inductance. Although resonance was maintained, charging power declined and so went the spark.

It was only a matter of time before the remaining jars gave out and a new capacitor constructed. It consisted of a glass plate stack with a final value of 0.027 mfd. While this value did not provide the intense spark produced with an 0.05 mfd. capacitor, it was adequate and did not stress the 40 year old coil (which had already sustained insulation failure). The first task, therefore, was to rewind the top two inches (approximately 10 turns). Evidently, this section had experienced a turn to turn breakdown as the result of the absence of a corona cap. Many sets of hands were required here so Wayne Nelson joined the four of us in the rewinding job. Insulating rope coated with Red Glyptol and a one turn corona ring of 3/4" copper tubing completed the winding task. In our absence, the Observatory's staff applied extra coats of Glyptol to ensure proper insulation properties.

Our next effort was directed to repairing or replacing the capacitor system. A test unit was made in which a 0.05 commercial unit was paralleled with the currently used glass plate stack for a total of 0.077 mfd. After several tests, John Foster felt that a value of 0.05 would be sufficient. His intuition was correct. It should be made known that John has constructed more Tesla coils than any of us and his vast knowledge has been augmented by a life long association with Mr. Strickfaden (see TCBA NEWS, Volume 2, #2, and Volume 3, #2).

It was agreed that the test capacitor of 0.05 would not hold up so another avenue had to be taken. Enter Ed Angell. Ed had constructed a large 0.14 mfd./50 KC oil immersed unit of the glass plate type. The oil served three functions: (a) electrical insulator, (b) prevention of the intrusion of moisture, and (c) of greatest significance as a coolant. Heat is one of the biggest problems faced with high power capacitors dealing with high frequency electric currents. Ed donated his unit to the project.

Before putting the unit into service, two considerations had to be looked at: (1) the location of the capacitor in the circuit and (2) its excessive capacitance. John pointed out that the voltages appearing across the coil or capacitor in a series resonant circuit are often many times larger than the charging voltage as supplied by the 60 Hz transformer. This is, indeed, especially true in high 0 systems. He felt it better to place the rotary gap across the transformer rather than the capacitor as had been the case in the past.

Since the capacitance of Ed's unit was nearly a factor of three, the unit would require some modifications. However, curiosity prevailed so before reducing the unit's capacitance, it was decided to try the outfit using the higher capacitance but with half the power for a brief interval. A brief interval it was as the 20-ampere breaker circuit shut itself down in a hurry. But the spark, OH WHAT A SPARK! So, off to the shop went the capacitor where Ed and the staff went to work dismantling and reducing its capacitance value. The closest that could be obtained was a 0.056 mfd.

We now knew that some form of current limiter had to be employed. Not far from the 60 Hz transformer lay a big 220 volt/6 KVA variac in unused form. This was put in series by the Observatory's electrician, Mr. Kim Johnson. To test the effect of our newly discovered variac, we had Mr. Johnson at its control (with handitalky in hand) and Wayne Nelson at the circuit breaker reading three A.C. ammeters (also with handitalky). We all were elated when the coil responded with a good looking display of sparks.

While all looked well, I suggested that the supply leads to the primary be connected from below. The reason for this is that it may have been the elevated connections that promoted a failure at the secondary's base during the 1940's. With the outfit now successfully working, the final task was the documentation of the parameters yet unmeasured. The system's low frequency numbers (voltage and current) as well as the primary capacitor's value had already been accurately measured. Values remaining were the coil's primary and secondary inductance, frequency and secondary output.

Because the primary and secondary inductance values were quite large, no inductance bridge could be located for their measurement. But, if the frequency could be accurately determined, then along with the measured capacity and known physical dimensions a reasonable inductance value could be ascertained. How could the frequency be determined? Enter Dr. Vic.

Victor Hymowitz is not only a high frequency Tesla coil enthusiast but a Ph.D. chemistry professor. He just happened to have a Tektronix 535A oscilloscope available for use. Dr. Vic put it to the test and came up with 121 KHz. With that data in hand, we went on to complete the determination of the unknown parameters.

When last seen in operation, the display had been improved by the staff at the Observatory. They replaced the glow tubes with a large multicolored neon wireless sign reading TESLA COIL (beautiful). In conclusion, we would like to express our thanks to Dr. E.C. Krupp, Mr. Wayne Nelson, Mr. Kim Johnson, Mr. Bob Brayden, and the entire Griffith Observatory staff for their extraordinary cooperation in this venture. Without their involvement, this project could not have been accomplished.

Coil Specifications

POWER INPUT- 3.5 KW being drawn from a 2 KW transformer for short intervals - Primary winding is connected for 220 volt but is supplied with 117 volts/32 amps drawn momentarily.

FREQUENCY-The Tesla coil operates at 120 KHz as measured by a Tektronix 535 oscilloscope.

PRIMARY COIL-4.5 of the outer turns is used. SECONDARY-292 turns (including 6 inner primary turns in series)

SPARK GAP-The spark gap is a rotary unit with 14 tungsten carbide tips turning at 1725 rpm (nonsynchronous).

VARIAC-A 6 KW/220 volt unit is connected in the low frequency circuit for current control.

SPARK-47 inches in length.