2002-11-10
Electrostatic Motor
This is my small implementation of Bill Beaty's soda bottle rotary electrostatic motor.
The rotor is a fluro light starter case, with three aluminium roof flashing plates stuck to it with epoxy. The plates and charge sprays are of the same flashing material. The rotor bearing is a normal graphite pencil. The whole assembly is tacked on top of a DSE zippy box with some blu-tack.
The power supply is a 10 kV surplus coronatron supply from a laser printer. It has two main bipolar EHT outputs, rated at +/- 4.7 kV (from the specs provided where I got it), but I found that into the 200M input of my HV probe it delivers -9.9 kV and +8.7 kV. So when the motor is not flashing over, and the corona load is low, the motor is probably running on closer to 20 kV. The motor will run from either rail separately, or across both, suggesting that, worst case, it will operate from 5-8 kV sources.
The images to the left and below are of flashing over runs, just for artistic effect. The motor runs quite a bit quicker when there is no loss from flash overs. It ran at about 200 RPM once I balanced the rotor carefully and added a piece of foam in the bottom of the rotor with a hole just larger than the pencil to stop it slapping around.
The sound created is quite unique. I'll try to capture some of it in the future. The corona from the charge sprays hisses loudly, and as the rotor rotates the edges and surfaces of the plates present the sprays with differing loads, chopping the corona hiss at three times the rotation speed.
Bill's design is targeted to very low supply currents and voltages. His can run on comb-run-through-hair type power levels. While mine being much smaller has large losses. Its plates are 'sharper' than his bottles and leak corona badly, even after I carefully smoothed their edges. My rotor is small and has little inertia which doesn't help, and the bearing is quite poor. During the testing the humidity was way too high for any kind of 'static electricity' testing, but considering the losses I doubt this little device will run from everyday sources.
Placing an insulator between the plates and the rotor makes no observable difference to the performance. Perhaps not surprising, the dielectric passes the electric field largely untouched. However, placing an insulator between the charge sprays and the rotor will slow or completely stop the motor. It appears that the ion current is critical to the operation, performing the job of the commutator in an magnetic motor.
Some carbon brushes, like the centre conductor of a spark-plug lead, should serve quite well as contact brushes instead of the sprays. This may lower the minimum voltage required to operate the motor, if the friction can be kept down.
The torque output wasn't measured. It didn't seem very much. I wasn't prepared to use my finger to load-up the rotor, as I would with an magnetic motor, and get a feel for its output!
At first I was a little concerned that the rotor would store charge like a capacitor and deliver shocks long after power down. Any capacitance it has must be very small, there is no such effect. Although it may be the edges of the plates leaking away the charge quickly. With larger devices this may become a problem?
As you can see in the dark shot, which was done with the plates very closely spaced, the main current flow is from the sprays as they charge and discharge the rotor plate capacitance. The top and bottom edges of the rotor plates seems to be the largest loss points on the rotor. The rounded corners of the stator plates obviously weren't rounded enough, bright spots of corona and associated strikes are visible.
The device seems to function more as a charge water-wheel than anything else. The rotor carrying charge from one ion spray to the other where it is neutralized and recharged in the opposite direction. The stator plates provide only the steady electric field for the rotor plates to work against. In this sense it is the electric dual of the magnetic motor, with the inductive components replaced with capacitive ones, and the current with potential.
Like a magnetic motor, a stronger stator field probably results in slower rotation but more torque. The analogy my old power engineering tutor used was 'like turning your arm in honey' which I always thought was cute. But there has to be a stator field there or nothing much will happen, however if the field is too low the unloaded motor can take off and destroy itself. The rotor is the same, it isn't current in the electric dual, it is voltage, so the higher the stator voltages the more torque, but higher voltages means more charge has to be sprayed onto the plates faster in order to charge their capacitance, so more current is pulled from the supply. Splitting the stator and spray supplies and doing some more controlled experiments is probably of value.
This thing should work as a generator! I can't see any reason why it wouldn't. It isn't very practical though, the polarising stator plates would need to be charged to bootstrap the process (same as a magnetic generator), and the output would be very high impedance. You would probably need to use contact brushes. There is a challenging future experiment; make an electrostatic generator out of this thing. You should be able to charge a capacitor via a very high resistance and flash a neon with it, or maybe crack a spark. If the polarising supply is separate, you will not need a huge resistor at all.
If contact brushes are used and lengthy efforts are taken to reduce the corona losses, rotor capacitance and friction this little device should run from a dry stack for years. Putting the entire device in a vacuum and using magnetic bearings would leave only the brush friction losses.
Leave a comment on this article.