Spark Gaps, Transformer Selection and the Art of Quenching

Part II

By: Richard Hull
7103 Hermitage Road
Richmond, Virginia 23228

There are numerous rotary configurations each with its own merits and limitations. Critical to rotary construction is a balanced wheel! A poorly built rotary can kill or maim as surely as a bullet. The motor powering the rotary should be a DC or universal type brush motor of high torque and have a top speed of at least 5000 RPM. A variable power supply or SCR/triac motor speed controller can be used to vary the speed of the motor. A large number of wheel contacts or points should be used with 6 being the minimum while 12 or more is best. The number of points on the wheel and its speed determine the number of power impulses per second that are being placed in the primary of the coil. It can be seen that this will control spark length and current within it. A second way of increasing the number of impulses per revolution is to place a second set of stationary electrodes at an alternate position around the wheel i.e., when one set of contacts is in line with sparking contacts, the other set is centered between a pair of contacts. This can double the power delivered to a coil at a given speed or allow you to retain the same power at half the former speed. Tesla did this at Colorado Springs.

Be sure to balance your input power/spark duty cycle/capacitor combinations because just continuing to pour more power or more interrupts per second into a small capacitance will not improve output. It will more likely reduce it!

The more contacts we place on a wheel the larger in diameter that wheel must be to allow spacing between contacts. At any given speed larger wheels mean more inertia, more powerful motors, more costly and stronger materials and finally there is more danger in operation due to inertia trying to pull the wheel apart or fling electrodes off into space!

The speed of a rotary can be critical since in higher impedance transformers RF energy is kicked back into the transformer system, thus causing the safety gaps to fire or even ruin potted transformer systems. The speed is often adjusted to “phase in” with the 60 hz. input power to completely eliminate safety gap firing. Needless to say, safety gaps across your transformer are a must. With large distribution transformers their adjustment can be tricky since when they do arc a virtual short exists. So stand by during the initial set up phase to kill the power.

Large transformers of the non-shunted type require ballasting or inductive current limiting in series with the primary input of the transformer. A couple of methods are used. First is ballasting, which involves IR or resistance limiting and can be accomplished with heaters (nichrome elements) placed in parallel with each other to reach the required wattage needed. In days of old, salt water rheostats were used but are messy and can take up valuable space. Resistance limiting is bad because it wastes power and generates lots of heat but has the advantage of being cheap and easy to construct. The input voltage and thus the output voltage of the transformer using resistance limiting is reduced as well as the current so a lot of “fiddling” with loads is necessary to get peak performance using this method. A couple of old coil hands have told me they tried this method but found it unsatisfactory and gave it up but also admitted they never really stuck with it long. I have used it with a 10kva distribution transformer and after 2 days of balancing, buying more space heaters and trimming with a variable toaster load I was getting excellent performance from a large coil. Plus, since it was winter, I stayed toasty warm to boot. I fed the system 220vac at 30 amps but calculate only half of this went to the coil; the rest went away as heat.

The inductive method of limiting current is the best but requires a giant variable inductor capable of handling the full operating current of the system transformer's primary circuit. The usual method is with a large high current Variac. Often, these items cost over $600 when new! Next comes a hand made inductor on a large form into which an iron core is inserted or removed to vary its inductance. These are bulky and heavy. Finally comes a coil wound on an iron core and tapped every so many turns and these taps are used to vary the input current to the main transformer. This last method is the easiest to construct and use. Hamfests often are a source of large Variacs and can be had in the 50-70 amp range for as little as $100. The Variac method is the neatest, most compact and smoothest to adjust of all methods mentioned. A coil can be quickly brought to peak efficiency using one of these babies and a change in system parameters such as more capacitance or gap change can be balanced or accounted for very easily at the turn of a knob.