W. Beaty 1997

Electrostatic design is something of a "black art" because there are numerous little-known rules which only become important at high voltage. Below are a few of them. Don't let these rules scare you off from building a device. Many projects work fine the first time. It's when they do not work that the following suggestions start becoming useful.

See also: Solving Humidity Problems and debug your VandeGraaff


Some types of static electric generators can only supply an extremely tiny current. The lower their current-generating capability, the more sensitive they are to stray leakage paths and the effects of humidity.

In the realm of electrostatics, a millionth of an ampere (uA, or microamp) is a fairly high current. For example, if you power an electrostatic motor with a plug-in DC high voltage supply having a 1000uA (1 milliampere) rating, you probably can violate all the following rules and ignore all the leakage issues. Where electrostatics is involved, 1 mA is a robust power supply with enormous output current. But if you try to use a Kelvin Waterdropper as a power supply, you'll probably have to follow all these rules in order prevent the surface leakage from shorting out your system. Also, if you follow all these rules, your device may tolerate much higher levels of humidity than otherwise.


Avoid using wood, cloth, cardboard, paper, masonite, or other fiberous materials as insulating structures. Their insulating properties will vary unexpectedly because of humidity changes, so on some days they are insulating, other days they become conducting. Stick with plastics and rubber. If possible, avoid glass as well. Glass surfaces tend to absorb a bit of moisture, and become conducting on really humid days. Plastic suffers less from this effect... but even the best insulator is not immune. On wet days, warm your plastic and rubber surfaces with a blow-dryer.


Sharp metal edges bad! Big rounded edges good! If a sharp-edged object is raised to a high voltage, tiny corona discharges will appear on the sharp parts and will silently and invisibly leak charge away into the air. The higher the voltage, the worse the problem. The closer the sharp edge is to an oppositely-charged object, the worse the problem. Also, generators with very low current capabilities, such as Kelvin Thunderstorm devices, are particularly sensitive to leakage, so sharp edges become a real issue for these. The problem can be partially relieved by keeping air away from the sharp edges. For example, split some tubing lengthwise with a razor, and use it to cover the edges. Even better, apply heavy beads of RTV silicone caulk to cover the exposed sharp edges. (RTV is the vinegar-smelling type. Avoid using water-based silicone.) A 1/4 inch thick layer is good.


Keep the metal parts far away from each other and away from "ground." Metal parts includes insulated wires. If oppositely charged parts are close together, or if charged parts are close to grounded parts, the e-field between them becomes extremely intense. This can cause sparks, but even worse, it can cause silent, invisible corona discharges to appear on the metal surfaces. The air between the parts becomes conductive, and the maximum voltage produced by your device will be drastically lowered. Support your connecting wires up off the table with insulating blocks. In general, the distance that causes through-air leakage will be large for large, blunt metal parts, and smaller for small parts, sharp edges, wires, etc. The distance where problems occur also gets smaller as voltages get higher. If you're trying to get to 100,000 volts, you should only be using large smooth spheres and cylinders, keeping them many inches apart, and covering any small protuberances and sharp edges with gobs of silicone caulk.


Keep the metal parts far away from each other. This includes insulated wires. If oppositely charged parts are close together, or if charged parts are close to grounded parts, they form a capacitor with significant value. This can slow the charge-up time of the device. For any particular current, the lower the capacitance, the faster the device charges to maximum, so you want to reduce the capacitance. The bigger the metal parts, the farther away they should be from each other.


I heard recently that ordinary vinyl record albums are poor insulators. I haven't check this out myself, but it does make sense. Electrostatic charging is a problem when cleaning records with a brush, and it causes them to attract dust for hours or days after being cleaned. Therefore, manufacturers might put some conductive chemical in the plastic. Perhaps some records lack this, and only some have problems. So, if you're thinking of building a Wimshurst machine, maybe it's a good idea to use some other type of plastic, and steer clear of old vinyl records for the Wimshurst machine disks.


Recently I was playing with metal objects suspended by nylon fishing line. Even in extremely humid conditions, the charge imbalance on the suspended object would not leak away quickly. It seems that the surface of thin fishing line is very small, and since humidity-leakage is across the surfaces of plastic, small surface leads to good insulation. So, if you are building some sort of static electric device, it might pay to suspend the conductive elements by using short lengths of fishing line. As with any device, keep the nylon lines clean by never touching them with greasy salty fingers.


Leakage current can flow through thin plastic insulation, so don't trust a wire's thin insulation to stop leakage. Instead support wires away from each other and away from conductive objects by using plastic supports and glue. You might consider using heavy solid wire, because once this wire is fastened securely at its ends, it can be sculpted to form any desired path directly through the air and away from all other conductive objects.


Avoid using extremely thin, bare wire. The thin wire qualifies as a "sharp edge", and corona discharges can appear. These leak charge away to the air. For voltage up to 10,000 volts, bare wire needs to be #12 or greater diameter, like coat-hanger wire. At higher voltages, bare wire must be replaced by copper tubing or pipe. Or, don't use *bare* wire at all, thin wire is OK if surrounded with very thick insulation.


Heavily insulated wire is available from some mail-order companies. Look for "high voltage" wire with ratings like 20KV or 50KV. If you cannot find a source, the next best thing is to use the center conductor of cable-TV cable, stripping off the black jacket and the copper braid. The center conductor will have a thick polyethelene or teflon insulation. Next best thing is "test lead wire," available in red and black colors from Radio Shack. Standard plastic-coated wire will usually work temporarily, but support it away from conductive objects such as tabletops. Another trick: give your wires a heavy insulation by sticking them through vinyl aquarium tubing.


If insulating parts become dirty or dusty, they can become slightly conductive, especially when humidity is high. Store your devices under a cover so they don't collect dust. And periodically clean any device which is used frequently. After all, even invisibly small amounts of corona discharge can emit ions which turn your device into an "electrostatic air cleaner," and it will attract all the dust, soot, and pollution out of the air and onto its insulating surfaces.


Those anti-static "dryer sheets" used to prevent static charging... they contain oil. Interesting fact: just a microscopic coating of oil will prevent "frictional charging." If you're rubbing fur on rubber, or passing a VandeGraaff belt over a plastic pulley, those surfaces need to be extremely EXTREMELY clean. The slightest bit of WD-40 can halt all charging. Just a few drops of spray coming from oily bearings can simulate the effects of 90% humid weather. If your "frictional" electrostatic machine stops working, yet the weather is dry, suspect that your insulating surfaces are no longer made of fur, rubber, plastic. They are made of oil. Fix: a good scrubbing in running water with some detergent (not soap!) will carry away the microscopic oil coating. Blot dry, then thoroughly dry the surfaces with a hair blow-dryer.


In general, the division between "insulator" and "conductor" is not absolute. Instead it's determined by the voltages and currents of the power supply and the circuit involved. Specifically: the line between insulator and conductor is drawn by the resistance of the whole circuit (or by the internal resistance of the power supply.)

For example, a flashlight might use 3 volts at 1 ampere, putting out three watts of visible and IR light. 3volts/1amp equals 3 ohms. So for a flashlight, any object with lots less than 3 ohms is a conductor, and any object with lots more than 3 ohms is an insulator. A flashlight "thinks" that a 1K resistor is a good insulator, while a 0.01 ohm wire is a good conductor.

For electrostatics, the numbers are quite different. A Kelvin waterdrop generator might produce 20,000volts at 1/2 microamp. Dividing this voltage by this current gives 40,000,000,000 ohms. That's the line between insulator and conductor. Quite different than three ohms! So, if an object is to act as an insulator, its resistance must be MUCH greater than forty billion ohms! For a Kelvin generator, a ten-megohm piece of wood will act like a very good conductor. Is it any wonder that a bit of surface moisture can convert an insulating object into a conductor? The highest value of resistor commonly available in catalogs is 33 megohms, and most electrostatic devices will see this device not as a resistor, but as a dead short.


With a robust electrostatic generator, you can test for proper functioning by lightly touching the metal parts and listening for a spark. But for barely-working devices, the sparks are far too small to hear or see. To cure this, place an AM radio within a few feet of your device and tune it to a blank station. The radio will pick up the electromagnetic pulses of even the tiniest sparks. Better yet, wear a "walkman" AM radio, and you turn yourself into a super-being, a Borg with enhanced sensory apparatus who can hear electro-magnetism. I've entertained myself by scuffing my feet on the carpet even in high humidity conditions, then touching grounded objects to make clicking sounds in the headphones.


Wouldn't it be a big help if you could *see* voltage while debugging your device? For example, you could adjust your VandeGraaff combs to produce maximum voltage on your machine. A simple trick: cut some short strips of tissue, then stick them to the metal parts of your device with a bit of tape at the top end of the strip, so that the strips hang down along the metal. When the metal becomes charged, the tissue strips will be repelled outwards. The further they rise, the higher the voltage on the metal object. Motion of the tissue makes the voltage "visible." (Note that this might not work in very dry conditions, since charge must leak along the tissue in order to give it an alike charge imbalance. Try painting your tissue with india ink and drying, in order to make it conductive.)


I discovered that India Ink makes a dandy conductor when dry. In cases where metal is expensive, you might consider using ink-coated wood or ink on papier-mache instead. Fancy shapes can be created in wood much more easily than in metal, then painted thickly with India Ink to guarantee a high conductivity. However, the conductivity of ink may not be high enough for use in spark-discharge electrodes. Metal should be used in this case. Try aluminum foil and rubber cement or contact glue, or find some adhesive aluminum foil tape (sold in rolls like duct tape, but with peel-off waxpaper covering the adhesive.)

LOOKING FOR BOOKS? Try searching amazon.com:

(try "science fair" too)


Created and maintained by Bill Beaty. Mail me at: .