Van de Graaff Generator Debugging

Don't Miss These:

FIRST, is your generator an off-the-shelf, non-kit product? If so, jump down to the section on FIXING YOUR COMMERCIAL GENERATOR



First see the article about SOLVING HUMIDITY PROBLEMS. VandeGraff generators need low humidity in order to function. Humid air causes insulating surfaces to become slightly damp, and they become a bit conductive. If you build a brand new generator and it won't work the very first time you run it, your generator might be bug-free, yet it will make no sparks at all. Before looking for other problem, always try the blow-dryer trick first. And use the balloon/armhair test to verify low humidity.


Oil and VDG machines don't mix. If the tiniest bit of oil gets on the rollers and belt, their surfaces will no longer touch. (Oil touching oil does not generate contact-electrification.) If you must lube your motor bearings, use great care and definitely DON'T use a WD-40 spray can under any circumstances. If you suspect that oil has coated your system, remove and clean both rollers and scrub down your belt with alcohol. Don't just wet and scrub, also use a bit of running alcohol, then rinse under running water. Your goal is to flush off the microscopically thin oil film, and if you don't flush with running water, you'll only stir the oil film around without actually removing it. If you're really paranoid you can even wear rubber gloves to eliminate contamination from skin oils. (At least scrub your hands thoroughly before starting the cleaning process.)

If one roller is made of felt, it might make more sense to replace the felt or the roller, rather than trying to clean out all the soaked-in bits of microscopic oil.)


What did you use for a belt? Beware, some types of black rubber are conductive and will not work. They're full of carbon, and will fail when installed in a VDG. They short everything out. Grey rubber or light tan rubber should be OK. To test your rubber for conductivity, simply rub a piece of it on your arm hair, then hold it near that hair and verify that the rubber surface attracts the hair and makes it rise. (Of course this test can only be performed when humidity is low!) If in doubt, avoid using black rubber.

Is your main support column made of insulating material? New and clean plastic water or sewer pipe is good. Wood is bad; it becomes electrically conductive when the weather is damp. Cardboard is similar: humidity makes it conduct. Assume that wood and cardboard are "resistor" materials, they make poor insulators unless they are baked in an oven and sealed to prevent water absorption. Stay with plastic instead. (If you have no choice, then bake your cardboard or wood until it is extremely dry, then coat it with layers of oil-based varnish.)

Roller materials are an issue. There should be no problem if one roller is polyethelene, and if the other roller is metal, and your belt is rubber. Some kinds of belt material and roller materials may not work. For your very first generator project, it is wise to use a reliable design. Once you have it working, you can experiment with other materials. If you have built a "powered" type of generator (having a DC hi-volt power supply inside,) then you can use metal for both rollers. For non-powered VDG machines, the surfaces of the rollers MUST be made from two different materials. (Think: if both roller surfaces were the same, then how could the generator "know" which end was (+) and which end was (-)? ) PS: if at least one roller is plastic, then you can experiment by placing other kinds of insulator on its surface. Try rubber sheets, teflon pipe-tape, paint or varnish, and verious kinds of adhesive tape.

In a contact-electrified or "frictional" type of generator (w/no HV supply inside), avoid making both rollers from metal even if you plan on coating one with plastic. A plastic-coated metal roller will probably form a sort of capacitor, and this would greatly reduce the roller's voltage. In a capacitor, opposite charges are attracted from the metal and then sit on the metal just under the dielectric surface. These opposite charges would vastly reduce the e-fields at the tips of the combs. So, if you intend to experiment with various roller materials, it's better to use two plastic rollers, then coat them with desired materials (tape, aluminum foil, etc.)


Are your combs positioned properly? The tips of the combs must not drag upon the moving belt. The tips of the combs should be within 1/4" of the surface of the belt. Hint for adjusting combs: short out your generator with a microamp meter, then set the comb positions for maximum microamperes (typically 5uA to 20uA). It's possible to make a microampere meter for a couple of bucks: an NE-2 neon bulb with a 0.01uF capacitor, and maybe a 1Meg resistor in series, for safety. The faster it flashes, the higher the microamps. (This from the old CL Stong, F. Lee Scientific American article on atom smasher" generating a half-megavolt electron beam with VDG and hard vacuum drift-tube.)

The combs must be properly connected. The upper comb must be electrically connected to the hollow sphere. The lower comb must be electrically connected to either the grounded case, or if you are building the "powered" style of generator rather than the "frictional" style, to the power-supply terminal


Does everything in your generator look OK, yet it will only produce tiny sparks? This problem can have several causes.

Normal electrode leakage: if you bring your fingertip near the generator, no long sparks will jump. This is normal. Your fingertip is too "sharp", and so it creates a corona discharge (tiny blue glowing spot seen in the dark). A corona discharge creates conductive air, which allows the charge to silently flow to your fingertip. To make longer sparks, bring the back of your wrist or the side of your arm near the main generator sphere. Or hold the bottom of a metal mixing bowl near the main sphere.

Sharp Edges: if your generator's main sphere has protrusions, then regions of corona discharge will appear upon the protrusions. The charge will not build up on the sphere, instead it will leak into the air. External protrusions such as screws, joints, edges of mixing bowls, etc., must be eliminated. Plastic protrusions are OK, it is usually only the sharp edges of metal which cause problems. Protrusions within the sphere are OK. The inside of your generator sphere can be a sharp-edged mess, it's only the outer surface which is important. (Keep any sharp-edged metal parts a couple of inches away from the belt and roller.) The outside of your generator's sphere should be smooth, and should have a donut-shaped area where the belt and support column enter the sphere. Quick and dirty fix: put thick layers of silicone caulk on any metal protrusions (thicker than 1cm.)

Tiny burrs: if a sharp metal part is exposed to strong e-fields, corona will appear, and electric charge will leak into the air. Therefor, make sure there are no sharp metal parts near the moving belt. If there are metal parts near the belt, they should be large and smooth, and have no sharp edges or tiny burrs. If in doubt, cover them with a thick coating of silicone caulk.

Corona jets: if the main sphere has some invisible corona leakage, sometimes you can find the trouble spot by searching for air jets. Tiny sparks of corona discharge will emit charged wind, and this charged wind feels ice cold. While your generator is operating, wave your hands slowly around the sphere while keeping them a few inches away from the metal. If you feel a jet of cold air, then you have found a source of leakage. My lips and face are more sensitive than my hands, so I move my face all around the sphere in order to locate the tiny sharp spots which emit cold, charged air.


First see the article about SOLVING HUMIDITY PROBLEMS. Commercial generators usually fail because of a combination of dirt and humidity. If the column, belt, or rollers are clean, then the generator might work ok even in fairly humid conditions. If they are dirty, then the generator might fail even if it is fairly dry out.

If you are certain that humidity is not your problem, verify that the belt and rollers move when the switch is turned on. You might hear the electric motor running, but does it actually spin the rollers and make the belt travel around the loop?

Verify that the "combs" are electrically conected properly. The lower comb must be connected to the metal of the base, and thence to ground. Or, if your generator uses a separate Hi-volt power supply within the base, then the lower comb must be properly connected to the power supply terminal. The upper comb must be electrically connected to the hollow sphere. The tips of the combs must within 1/4 inch of the belt. They should not drag on the belt; they do not need to touch it at all.

And as I mentioned above, a microamp meter is very handy for tuning up the belt/rollers/combs assembly. Use a microamp meter to short out your generator by connecting the meter leads to the sphere and to the metal base (or to ground.) Run the generator and the meter should indicate a few microamps. Do anything you can to make this value higher (tweek the comb positions, remove and clean the belt, dry the parts with blow-dryer, etc.) This improves the charging current but it cannot fix high-voltage problems caused by filth on the column or dust-speck corona-points attracted to the main sphere.

If your generator still does not work, try more of the fixes discussed at the top of the page. In particular see the part about oil contamination.

Created and maintained by Bill Beaty. Mail me at: .
View My Stats