here is a real pulse generator

Theory of Operation Results EM Field Theory and Wire Fragmentation? Isn't Defacing Money a Federal Crime? So Who Invented this Crazy Device? References

Theory of Operation: The Quarter Shrinker uses a technology called high velocity electromagnetic forming, or "Magneforming". This is a "high energy rate" process that was originally developed by the aerospace industry in conjunction with NASA, and was popularized by Aerovox, Grumman, and Maxwell. It involves quickly discharging a high energy capacitor bank through a work coil to generate an extremely powerful, rapidly changing magnetic field which then "forms" the metal to be fabricated. The technique uses pulsed power to generate a very high current pulse over a very short time interval. Although electromagnetic forming works best with metals that have relatively high electrical conductivity (such as copper, silver, or aluminum), it will also work to a limited extent with poorer conducting metals or alloys such as steel or nickel.

To shrink coins, I charge up a large high voltage capacitor bank consisting of a number of large "energy discharge" capacitors. Each capacitor is specially designed to reliably store up to 12,000 volts and deliver 100,000 ampere discharges. Each steel-cased capacitor measures 30" x 14" x 8", and weighs 165 pounds. A High Voltage Double Pole Double Throw (DPDT) relay is used to connect the capacitor bank to either a high voltage DC charging supply, or to a bank of high power "bleeder" resistors. A 15,000 volt high voltage transformer and a set of 40 kV rectifiers make up the DC charging supply for the capacitor bank. The electrical energy stored in the capacitor bank is proportional to the square of the stored voltage, and the actual "shrinking" force is proportional to the energy stored in the capacitor bank.

During shrinking, the charged capacitor bank is quickly discharged through a single layer work coil made from two-layer film-insulated 200C magnet wire. The coin is held firmly in the center of the coil by a pair of insulated dowel rods so that its axis of rotation is parallel to the centerline of the coil. This helps to keep the coin from twisting, and also helps balance axial forces that might otherwise eject the coin from the coil. The two ends of the coil are stripped of insulation, formed appropriately, and then firmly bolted to a pair of heavy copper bus bars. A spark gap is the only affordable device that can hold off the high voltage and then reliably and efficiently switch the huge currents involved in the shrinking process (typically 50,000 - 100,000 amperes). Originally, the high voltage "switch" that discharged the capacitor bank into the work coil was a special type of high power triggerable spark gap called a "trigatron". The trigatron switch was "fired" by applying a high voltage (~50 kV) triggering pulse to the trigger electrode, which in turn caused the main gap in the trigatron to ionize and fire. However, I needed to expand the working voltage range and reduce spark gap maintenance, so I have converted to a solenoid-driven high current spark gap switch that uses 2.5" diameter brass electrodes (similar to those used in the earlier trigatron switch). The solenoid drives one electrodes close to the other, triggering an arc. However, the movable electrode does not quite contact the fixed electrode in order to prevent contact welding (a potential problem at lower power levels).

When the spark gap fires, current rapidly climbs in the work coil, and the rate of change may approach five billion amperes/second. As the work coil current increases, it creates a rapidly increasing magnetic field within the work coil. The natural resonant frequency of the LC circuit that's formed by the capacitor bank and the inductance of the work coil ranges between 8 to 12 kilohertz (kHz). Through electromagnetic induction ("transformer action"), a huge circulating current is induced within the coin. However, because of skin effect, most of the induced current is confined to the outermost rim of the coin, typically penetrating to a depth of only 0.050 inches or less. In clad coins, most of the circulating current actually flows within the better conducting copper center of the clad sandwich than in the outer layers. This causes the center of clad coins to shrink a bit more than the outermost layers, leading to an "Oreo Cookie" effect. Because of Lenz's Law, the magnetic fields of the coin and work coil strongly oppose each other, resulting in tremendous repulsion forces between the work coil and the rim of the coin. The circulating current in the rim of the coin prevents most of the magnetic field from the work coil from penetrating the interior of the coin. The initial energy stored within the capacitor bank is typically in the range of 3,500 - 6,300 Joules (watt-seconds). Because this energy is discharged in as little as 20 millionths of a second, the instantaneous power is very large and, for a brief instant, is roughly equivalent to the electrical power consumed by a good sized city. The repulsion forces between the work coil and the coin create tremendous radial compressive forces that easily overcome the yield strength of the metal alloys in the coin, causing the coin to plastically deform into a smaller diameter. The higher the initial energy, the greater the degree of "shrinkage". Applying a 6,300 joule pulse results in a quarter whose final diameter is about 0.1" SMALLER than a dime! At the same time, similar radial expansion forces cause the work coil to explode in a potentially lethal shower of copper fragments. In addition, there are also large (but less strong) axial forces that squeeze the work coil wires together while the coil is simultaneously expanding in diameter. In all cases, the forces acting upon the coil are in a direction that would tend to increase its inductance. The coin effectively behaves as a short circuited single turn secondary in a 10:1 step down transformer, and the current that's induced in the outer rim of the coin may exceed a million amperes! The metal forming effect of huge magnetic fields is sometimes seen on a much larger scale. For example, repulsive forces between the windings of large utility power transformers can literally tear the transformer apart during a high current short circuit, or rip heavy bus bars from their mounting insulators within electrical substations.

As the work coil expands, the insulation separates from the wire (since the film insulation can't stretch as much as the ductile copper!), the wire eventually fragments, and pieces of the coil are forcefully ejected outward with the force of a small bomb, with small coil fragments reaching velocities of up to 5,000 fps. For safety, the work coil is housed inside a blast shield made from 1/2" Lexan polycarbonate, the same material used to make bulletproof windows. Furthermore, the regions in the direct path of the exploding coil fragments are further reinforced with 1/4" thick steel plates. Once the work coil disintegrates, most of the residual energy in the system is dissipated in a ball of blue white plasma. The Quarter Shrinker is designed so that any residual energy in the capacitors is safely dissipated by high power bleeder resistors. The system is triggered from about 15 feet away from a remote control box. I've found (the hard way!) that 8,000 Joules is about the maximum energy I can repeatedly use without running a risk of fracturing the Lexan blast shield from the shock wave. Under the right conditions, Lexan does shatter - I've got the pieces to prove it! Other experimenters (Rob Stephens, Bill Emery, Phillip Rembold, Ross Overstreet, Brian Basura, and Ed Wingate) have resorted to using steel enclosures when running at higher power levels. Adding 1/4" steel plates has stopped the Lexan blast shield from fracturing. However, future designs will use 1/2" thick AR400 steel armor plate to better withstand deformation from repetitive hammering by supersonic coil fragments.

In 2009, the folks at Hackerbot Labs (Seattle, WA) built their own coin shrinker. By using a 100,000 frame/second camera, clear Plexiglas dowels, and carefully pretriggered electronic flash units, their partners at Intellectual Ventures, Inc. were able to actually capture the coin AS IT WAS SHRINKING. Because the shrinking process occurred so rapidly, "shrinking" is only seen during four consecutive frames (or about 40 millionths of a second for their larger capacitor bank).

Results: The largest coin I've shrunk was a Silver Eagle, a silver coin that starts out being about 1.6" in diameter, and ends up 1.3" in diameter afterwards. At 6,300 joules, a silver Morgan Dollar is reduced from about 1.5" to 1.25" in diameter, and a clad Kennedy half dollar is reduced to a diameter smaller than a US Quarter. At 5,000 joules, quarters will shrink to about 0.080"-0.100" smaller than a dime. During the shrinking process, the coil has recently found to fail just after the first current peak. Fortunately, virtually all of the coin shrinkage occurs during this time. Disintegration of the coil prevents the energy discharge capacitors from seeing voltage reversals that can potentially damage them. However, the rapid discharge and tremendous surge currents are still very hard on most capacitors. Because of premature failures with earlier GE pulse capacitors, I've redesigned the system to use low inductance, 100 kA/shot Maxwell (now General Atomics Energy Products - GAEP) pulse capacitors that are truly rated for this type of abuse (300,000 shots at 100,000 amperes/shot). My original capacitors would begin failing after 50 - 100 shots. The more robust Maxwell capacitors have withstood well over 6,000 shots with nary a whimper.

Examination of the coil fragments show that the wire has been substantially stretched (#10 AWG looks like #14 AWG afterwards), it becomes strongly work hardened, and has periodically "pinched" regions and kinks caused by the copper being stressed beyond its yield strength by the ultrastrong magnetic field. Many fragments are 1/4" long or less, and all pieces show evidence of tensile fracture at the ends. Since the wire's insulation is blown off, most fragments are bare copper. The wire often also shows signs of localized melting on the inner surface of the solenoid due to current "bunching" from a combination of skin effect and proximity effect.

A larger diameter work coil, operating at lower power levels, is used to crush aluminum cans. An aluminum soft drink can ends up looking like an hourglass as the center is shrunk to about half its original diameter. In this case, the coil does not disintegrate due to its more massive design (3 turns of #4 AWG solid wire) and the system is fired using a lower energy level than that used for coin crushing. At higher power levels the can is ripped apart from the combination of the air inside the can suddenly being compressed, and the heating of the can from the induced currents. Can crushing also works with steel cans, but the can undergoes greater heating and reduced shrinkage because of steel's lower conductivity. The "skin depth" is also much thinner due to the ferromagnetic properties of the steel alloy. Since the work coil is not destroyed during can crushing, the capacitor bank and spark gap are stressed by a damped oscillatory ("ringing") discharge. The capacitor bank voltage must be reduced to so that oscillatory voltage reversals don't overstress and damage the pulse capacitors. Since most of the capacitor bank's energy ends up being dissipated as heat in the spark gap, can crushing causes also significant electrode heating and erosion of the electrodes in the HV switch.

The Quarter Shrinker works very well on clad dimes, quarters, half dollars, Eisenhowers, silver Morgan and Peace Dollars, Susan B. Anthony, Sacagawea, small Presidential dollars, and most foreign coins. It works less well with nickel and nickel-copper coins, but only slightly works with plated steel coins. It also works well with older bronze and copper-zinc alloy pennies. However, since mid 1982, US pennies have been made using a zinc core with a thin copper overcoat. During shrinking, the thin copper layer vaporizes and the zinc core melts, leaving an unrecognizable disk of molten zinc accompanied by a messy shower of zinc globules throughout the blast chamber. Because of the greater hardness and much poorer electrical conductivity of nickel-copper alloys, the shrinking process doesn't work as well with US nickels, only shrinking them by only about 10% even at 6,300 Joules.

A shrunken coin weighs exactly the same as before, and its density also remains unchanged. The coin becomes thicker as its diameter is reduced, but the overall volume of the coins stays the same. Certain bimetal foreign coins (with rings and centers made from different alloys) may show different degrees of shrinkage based upon the electrical conductivity and hardness. In some cases, the center portion may shrink a bit more, freeing it from the outer ring. This occurs with older Mexican, UK, and French bimetals, and newer Two Euro bimetal coins.

Because of the extremely high discharge currents and fast current rise times, energy discharge capacitors are fabricated to have low inductance and use special internal construction techniques to safely handle the mechanical shock created by magnetic and dielectric forces during high current pulse discharges. Unfortunately, the GE pulse capacitors that I previously used were simply not designed to withstand this abuse, and they began to fail catastrophically. One actually ruptured its metal case, hemorrhaging stinky, arc-blackened capacitor oil all over the floor. This was a real hit with the wife! Our Maxwell energy discharge capacitors have proven to be true "Timex's" - they "take a lickin' and keep on tickin'".