Debugging Van De Graaff generator

[Artlav]

I'm trying to build a simple Van De Graaff generator, for the first time.
The problem is quite obvious - the current generated is zero.
Can somebody help to find what is wrong?

The column is a typical sewer pipe, axles are aluminum bars, electric screwdriver acts for motor.

The upper roller is glass textolite in free-spin on aluminum bar, the belt is doorframe rubber band. The materials seems to fit the purpose - when the roller was rubbed against the belt, it got enough charge to lift some pieces of paper.

The lower roller is a duct tape construct fixed on an aluminum bar the screwdriver is connected to.

Brush wires are typical copper wires, the sphere is a party balloon rolled in a layer of aluminum foil and a layer of duct tape, also tried a dried pepsi can and two pots.

The weather is bad, but the attempts are made in an air-conditioned room.

There are neither sparks, nor current indication on an ampermeter, with or without the sphere.

The pictures of the thing:

Overview:
http://orbides.1gb.ru/orbf/vdgdbg_1.jpg
Upper brush:
http://orbides.1gb.ru/orbf/vdgdbg_2.jpg
http://orbides.1gb.ru/orbf/vdgdbg_3.jpg
http://orbides.1gb.ru/orbf/vdgdbg_4.jpg
Lower brush:
http://orbides.1gb.ru/orbf/vdgdbg_5.jpg

Any advice where to look for problems, or is anything obviously wrong?

 

[dlgoff]

The copper wire brushes need to be on/touching the rubber bands inorder to transfer the electrons. Do you have the motor-end brushes Earth grounded? Also, why do you have the sphere so far away from the anode brushes? You might be getting leakage on the wire leading to it.

Now don't expect too much current. I would only expect a few microamps. Do you have a meter to measure at high voltage dc?

 

[Manchot]

I have some experience with this: I'm an EE grad student who built a VDG in high school. As dlgoff said, I don't think that you're going to measure any current with a typical ammeter. However, in my experience, you do not want the brushes touching the belt: you want them to be facing the belt so that charge can arc across the air gap. Otherwise, your belt will wear out from the friction. You want them to be as close as possible without touching.

The biggest problem, of course, is that your ball is essentially doing nothing as it stands. The entire point of a VDG is that you are supposed to be moving charge up a electrostatic potential through mechanical means. All you are doing is attempting to dump charge onto a sphere from the outside. After a very small amount of charge is deposited, a field will build up preventing the brush from collecting anything else. To be honest, you may as well not even have the sphere.

What you want to do is make it so that the brush is inside the sphere. Because there is no field present inside a conductor, you can keep depositing charge for as long as your motor is able to push it up the field seen on the outside of the sphere. Also, a foil sphere might work a little bit, but remember that any sharp edge will allow charge to leak off. (Sharp edges produce high fields, allowing the air to ionize. This is how lightning rods work.) Ideally, you want your sphere to be as smooth as possible. You might try taping together two metallic bowls, or you might buy one of those decorative lawn balls and cut a hole in it for your column. (You can then coat the sharp edges with something nonconductive, like electrical tape or rubber.)

Here's a site that I used as a reference when I built mine. Be careful: there are some kooky things on there, so if I were you, I'd stick to the VDG portion.
http://www.amasci.com/emotor/vdg.html

 

[Manchot]

Oh yeah, I forgot to mention: don't use an ammeter to measure the current. Not only will it not work, the high voltages might just destroy the ammeter's electronics.

 

[russ_watters]

How would you even use an ammeter for this? You need it wired in series.

 

[Artlav]

Here's a site that I used as a reference when I built mine. Be careful: there are some kooky things on there, so if I were you, I'd stick to the VDG portion.
http://www.amasci.com/emotor/vdg.html

Thank you for the link, that cleared some questions.

The biggest problem, of course, is that your ball is essentially doing nothing as it stands.
...
What you want to do is make it so that the brush is inside the sphere.
...
Also, a foil sphere might work a little bit, but remember that any sharp edge will allow charge to leak off.

That seems to be the main problem - with the foil balloon it didn't work in any conceivable configuration and any available pair of material.

Replacing the balloon with a metallic pot of a same diameter as the pipe gave an immediate result as some slight sparking, then replacing the door insulator belt with martens bandage rubber band made the sparks visible and up to a centimeter long.
Thank you for the advice again.

Now the question is - what should be the optimal performance on a structure of that size and design? It feels rather weak now, and I've seen much smaller ones with much more power.

Oh yeah, I forgot to mention: don't use an ammeter to measure the current. Not only will it not work, the high voltages might just destroy the ammeter's electronics.

Hm. What can i use, then? The amasci site advices to attach an ampermeter to the device to have a value for tuning, and the multimeter is the only thing precise enough that i have.
Will a classic coil-needle ampermeter work, or it needs something special?

The copper wire brushes need to be on/touching the rubber bands inorder to transfer the electrons.

However, in my experience, you do not want the brushes touching the belt: you want them to be facing the belt so that charge can arc across the air gap. Otherwise, your belt will wear out from the friction. You want them to be as close as possible without touching.

So, is there a definitive explanation on what should the brush positions be?

The ones i came around range from quarter inch to contact, facing to perpendicular and with reasons being - no electricity due to contact short-out, gap needed for charge transfer, the belt will wear out, etc.

What is the optimal distance/orientation, and why?

 

[dlgoff]

If the brushes don't rub against the bands, how are the going to move the electrons from the sphere to ground? I worked on a 2MeV van de graaff in college (accelerated protons and He++ with beam currents in the microvolt range). It had a belt about one foot wide with the brushes covering the entire width. This was to get the maximum electron transfer. Your rubber bands will ware out but you won't get any voltage on the sphere unless you rub them.

 

[Manchot]

If the brushes don't rub against the bands, how are the going to move the electrons from the sphere to ground? I worked on a 2MeV van de graaff in college (accelerated protons and He++ with beam currents in the microvolt range). It had a belt about one foot wide with the brushes covering the entire width. This was to get the maximum electron transfer. Your rubber bands will ware out but you won't get any voltage on the sphere unless you rub them.

If your belt is good enough, the electrons will arc across the air gap from the belt to the brush. The fact that the brush is made of pointy ends promotes ionization.

 

[dlgoff]

Sure sounds like a waste of energy to me. Besides, we're only talking about rubber bands here.
Edit: Okay. You don't have to have the brushes touching the belt as you say. But I think in this case they should.