Improved Can Crusher
As you may recall, my last attempt at this was
rather pitiful. In fact, it only succeeded in frying about $50 worth
of high voltage diodes. To redeem myself, I first went on a knowledge
quest, reading patents and web pages (see, for example, U.S. Patents: 4,198,590
"High Current Triggered Spark Gap"; 4,517,497
"Capacitor Discharge Apparatus"; 5,585,696
"High Current Density Glow Discharge Switch"; 5,850,125
"Pseudospark Switch Having an Insulator Between Electrodes"; 5,126,638
"Coaxial Pseudospark Discharge Switch"; 4,939,418
"Gas Mixture for Triggerable Spark Gaps"; 6,033,565
"System for Treating Gases or Fluids With Pulsed Corona Discharges")
dealing with the subject of high power pulses. Ignorant no more,
I devised several improvements to the old system. First, I installed
my high voltage RFI filter (normally used for
my Tesla coil) between the neon sign transformer
and my Cockcroft-Walton voltage multiplier.
After the C-W, I added a high voltage current
limiter. This was followed by a high voltage
relay which allows me to break the capacitor charging circuit and redirect
the power to the switching electrode of a triggered
spark gap. The following is a block diagram of the new system.
The crusher itself is simply a piece of 1/2" copper tubing formed into
a two turn coil wrapped around a 12" length of 3" OD polycarbonate tubing.
It is supported by 1/2-13 fiberglass threaded rods mounted on a piece of
1/2" thick polycarbonate sheet. The capacitor is actually two oil
filled GE capacitors in series. Each is rated 3 microFarads at 20,000
volts. The measured capacitance is 1.46 microFarads. At 40,000
volts, they should store a total of 1,170 Joules of energy. The coil
a measured inductance of 0.3 microHenrys. When the gap fires, the
resulting LC oscillator should have a natural frequency of 240,000 Hz.
Thus, rise time of the current pulse should be roughly one microsecond
(1/4 cycle), giving an instantaneous power level of 1,120,000,000 watts.
The following picture shows the business end of the system, ready to run.
Respectful of the power levels involved, I decided to start the runs
at only 20,000 volts. First, I set the main gap of the triggered
spark gap to fire at 24,000 volts. Then I charged the capacitor bank
to 20,000 volts and energized the relay. A bright flash and muted
snap from the gap suggested that it had functioned as I had hoped.
A quick check of the capacitors with a high voltage probe showed that they
had indeed been discharged. There was no effect on the can.
The gap was widened, the capacitors charged to 30,000 volts, the relay
tripped, flash, snap same result. 40,000 volts, everything I had,
still nothing. I began to suspect that the problem was with the coil.
With such low inductance, it must not have been generating much of a magnetic
field. Time to improvise. The next picture shows the setup
with a coil made from ten turns of insulated 14 AWG stranded copper wire
with some copper sheet tabs soldered to the ends. It is held to the
tube with several strips of PVC electrical tape.
Since the gap was already set for operation to 40,000 volts, I charged
the capacitors all the way and let fly. The first notable difference
from the earlier runs was a very loud and deep bang. A quick glance
at the can convinced me that this sound did not come from the spark gap,
but rather from the air being expelled from the collapsing can at supersonic
speed. Also, since only the top of the can was crushed, I assumed
that it must not have been well centered within the coil, so I prepared
another can (by punching a 1/2" hole in the bottom, to balance the air
rushing out, so as not to have a jet propelled can [while possibly amusing
in some circumstances, it seemed like a bad idea in my small basement]),
centered it carefully and let it have it. The results are seen in
this next picture.
For those who might be interested, this coil has an inductance
of 9.8 microHenrys. The natural resonance frequency is now 42 KHz,
so the power level is only 200,000,000 Watts, with a peak current of 5,000
amps. Now, you may think that this was a lot of effort to go through
just to crush a pop can. Your right, but I have further
plans, stay tuned.
A bit of a post script here. After several runs with the can crusher,
I noticed that the steel spheres in the triggered spark gap had turned
brown. The only explanation I can come up with for this is the known
photolytic instability of PVC. The PVC destined for exterior use
(siding, shingles, etc.) is normally blended with ultraviolet stabilizers
(kinda like sun block) to slow down this process. I suspect that
the UV radiation (maybe some X-rays too) from the firing gap is inducing
dehydrohalogenation in the PVC, generating hydrogen chloride. The
hydrogen chloride would then react with the water vapor inside of the gap
case to generate hydrochloric acid, which would rapidly corrode the steel.
The picture below shows the result. One ball has been cleaned with
an extra-fine sanding pad for comparison (it cleaned up fine, no pitting,
just a uniform brown coating).