High Voltage Please, But don’t Forget the Current

In high voltage applications involving tens of thousands of volts, too often people think about the high voltage needed but don’t consider the current. This is especially so when part of the circuit that the charge travels through is an air gap, and the charge is in the form of ions. That’s a far cry from electrons flowing in copper wire or moving through resistors.

Consider the lifter. The lifter is a fun, lightweight flying machine. It consists of a thin wire and an aluminum foil skirt separated by an air gap. Apply 25kV volts across that air gap and it lifts into the air.

So you’d think that the small handheld Van de Graaff generator pictured below, that’s capable of 80kV, could power the lifter. However, like many high voltage applications, the lifter works by ionizing air, in this case ionizing air surrounding the thin wire resulting in a bluish corona. That sets off a chain of events that produces a downward flowing jet of air, commonly called ion wind, lifting the lifter upward.

But that ionization of air requires current, electrons flowing, in the wires coming from a high voltage power supply. In fact, that ion-filled air gap is the equivalent of a high resistance wire in the circuit, along with some capacitance; it’s a part of the circuit.

Lifter circuit and equivalent circuit with resistor and capacitor instead of the lifter
Lifter circuit and equivalent circuit

A Van de Graaff generator, even a DIY 84kV one, is a low current power source and cannot supply a high enough current to ionize enough air fast enough to produce the necessary strong jet. However, the power supply powering the lifter above converts the energy from a wall socket using a flyback transformer and a Cockcroft-Walton voltage multiplier. That way it produces sufficient voltage and supplies more than enough current.

Another example is the corona motor, a type of electrostatic motor, that also works by producing a high voltage across an air gap, multiple air gaps in fact. The gaps are between sharp metal blades and a neutral plastic cylinder. The blade sprays ions across the gap onto the cylinder.

Corona motor powered by the triboelectric effect
Corona motor powered by the triboelectric effect

However, the gap width is very short, requiring a lower voltage. And more importantly, the cylinder doesn’t have to be ionized much for the cylinder to start turning meaning that not as many ions are needed. i.e. the required current is lower than with the lifter. In this case perhaps the lowest current high voltage power source I’ve ever worked with, rubbing a PVC pipe with a cotton cloth, is sufficient. That utilizes the triboelectric effect.

And then there are applications where all that’s needed is to accumulate charge in a capacitor until there’s enough, taking as much time as necessary. An obvious example is to simply produce a big spark.

An example of that is a TEA laser. A TEA laser works by accumulating charge across a small spark gap and in two flat plate capacitors. When the voltage has built up sufficiently to breakdown the air in the spark gap, it fires, causing a subsequent sparking across the lasing channel, resulting in a laser beam. In the photos below you can see the laser being powered by a low current Wimshurst machine and by the same powerful Cockcroft-Walton power supply mentioned above for flying the lifter. Both produce an identical voltage, accumulate the same amount of charge and fire an identical laser beam. However, the Wimshurst machine requires around 12 seconds of hand-cranking to do so, resulting in the laser firing only every 12 seconds. The Cockcroft-Walton power supply fires the laser around every 1 second.

How do you choose a high voltage source with sufficient current to match your application? This is largely done by experience. The lifter needs to move a large mass of air, and to do so a large quantity of ions are required, and hence a high current to create those abundant ions. On the other hand, a corona motor works using the Coulomb force. Blades of one polarity repel areas of cylinders that are charged with the same polarity. With just a fan as the load on the cylinder, not much repulsion force is required. Naturally for a bigger load, more force would be required and hence a faster charging rate and so more current. Similarly with the TEA laser, more frequent laser beams require higher the current.

I’ve already taken you through a range of high voltage sources available to you. With the needed high voltage already available from any of those sources, we now need to look at the current available from them.

The flyback transformer and Cockcroft-Walton voltage multiplier power supply gets its power from the wall socket. Taking into account losses in the various resistors, transistors, capacitors, diodes and the flyback transformer itself, there’s still a relatively large amount of current available, even if it is in the single- or double-digit milliamps. For high voltage that’s considered quite a bit (remember that power is the product of voltage and current).

On the other end of the scale, the triboelectric effect works by a transfer of electrons when making contact between two specific materials, and the retention of those electrons when contact is broken; basically, it works by rubbing the materials together. In that case, very little charge is transferred compared to the current coming from a wall socket.

The Van de Graaff generator actually starts with the triboelectric effect. It’s the making and breaking of contact between the rollers and the belt that is the rubbing together of two materials. However, unlike rubbing a cotton cloth against a PVC pipe by hand, in a Van de Graaff generator the rollers can be rotating at hundreds of RPM, generating charge more rapidly. But the amount is somewhere in the low microamps for a tabletop Van de Graff, small compared to wall socket current (that’s assuming you’re taking charge directly from the Van de Graaff’s dome and not waiting for sparks.)

And a Wimshurst machine generates its charge by induction when sectors on the opposing disks pass a neutralizer bar. Surprisingly, it’s possible to dimly light a small 20mA LED, where the LED is placed in series with the neutralizer bar. However, by the time the charge is inefficiently removed at the collectors, it’s reduced substantially. The current from the output of a Wimshurst machine is usually in the single digit microamps (again, we’re not waiting for sparks.)

Those are some of the examples I can recall where people, myself included, have forgotten that just because high voltage is involved, that doesn’t mean that basic electronics no longer applies. I’m curious what examples you’ve encountered, either where you or others have forgotten about the current or maybe even some other electrical property. Let us know in the comments below.

28 thoughts on “High Voltage Please, But don’t Forget the Current

  1. Nice writeup. Another source of information on EHD propulsion here http://www.blazelabs.com/l-intro.asp

    And interestingly enough MIT has looked at it as a legitimate propulsion for aircraft, they claim 110 newtons of thrust per kilowatt, compared with a jet engine’s 2 newtons. An article here http://lae.mit.edu/ehd/

    I am working on designs and energy storage systems for a RC scale build, the main problem is getting the power supply and storage system airborne requiring alot of current and a diminishing return on mass. New energy storage systems are required but will come in time.

    If anyone wants to call me a quack or help you can find me at HAD.io

    1. Yeah, I remember when Savior (blazelabs.com) achieved his 100 gram payload lifter, a record that to my knowledge hasn’t been beaten.
      Regarding ones that lift their own power supply, I keep an eye out for that tech too. The last improvement in tech toward that that I know of was the supercapacitor. Still needs the electronics to step up the voltage, but at least with a supercapacitor there’s enough energy storage for a few seconds of lift.

    2. I remember looking to see if that had been attempted. There were quite a few interesting ideas out there. :)
      I don’t know how to math-it-out but there must be a certain scale at which RC lifters could work. I want it to be possible. Maybe a HV plane; not vertical take-off?

      The HV supply, battery, and wires seem to be the bottlenecks.

      I want to see a blimp that uses ‘lifter tech’ to maneuver itself. I think that would definitely be doable. It could be toroidal-shaped. :D

      You could make the lightest RC lifter possible that you can and possibly fly it inside a clear cage with a heavy, non-flammable gas.

    1. I do agree with you for the most part, but it’s difficult for kids to understand why 12,000 volts of socks-on-carpet is safe but 120 volts from the outlet is not. I feel that after about 3rd grade they should definitely be taught right.

      I still can’t get my father to understand what 3 phase means. O_o

      1. Although I appreciate the concept is difficult to teach. If current was taught along side voltage the dangers are more apparent. Too much focus is on voltage, an analogy with water reveals the dangers.

        It took me years to realise that static electricity was the same as “normal” electricity.

        The media doesn’t help with scare stories about High Voltage, I’ve never seen a headline referencing high current.

        3 Phase, oooh real voodoo :)

  2. Forgotten: Lord Kelvin’s water droplet HV generator… And wasn’t there a crystal, which when heated (from very low temp) that produced MVolts.. also Piezo transformers which are overlooked, but are really nice generators too.

  3. Notable absence from all these HV sources is good old automotive ignition coils – modern ones can be 5v triggered, internally controlled (current limiting etc) and generate huge power compared to the systems of yesteryear.

    I’ve often wondered about hooking one up to the mains via a transformer to power a jacob’s ladder, would save the need for any sort of driver circuit.

    1. I thought about including my 555/ timer chip/ignition coil power supply in the list but I have that and so many more that I had to draw the line. Since I built it and used it only once, I left it out. Do the modern ones you mention have additional circuitry inside? Sounds like they do.

      1. Yeah, as things moved from distributor to coil-on-plug under ECU control they quickly worked out that putting the coil driver circuit inside the ECU was a massive PITA for EMC & reliability, so most modern coils have a driver inside them (have a look at Bosch Bip373, or VB921 devices for a rough idea of the features), you give them +12v, GND, and trigger them with a logic level signal.

        There are external driver devices, Ford’s EDIS system being one I’m most familiar with, very robust and a brilliant bit of engineering, but the all-in-the-coil flavour is undoubtedly the neatest solution.

    2. I’ve never used any kind of driver circuit with a Jacob’s ladder, I just hooked the output of a sign or burner transformer directly to the vertical rods of the ladder.
      It works well, but only with transformers that don’t mind producing continuous arcs.

    3. I shouldn’t making a comment like this when I tired and not feeling well, and am killing time between the periods when my body lets me rest. There may be errors. Keep in mind an ignition coil isn’t a typical AC transformer. The high voltage appears at the secondary when the current in the primary is cut. With an 8 cylinder engine operating at 5000 RPM that happens 40,000 times per minute. In the US the line current is zero 7,200 each minute. For an ignition coil to create a fat spark the the electromagnetic field needs to saturate the coil for amount of time and the current needs to be sharply cut at maximum saturation. The line AC frequency my be outside the design parameter of an ignition coil. For kick rectify the line into a square wave using rectifiers rated for the peak line voltage, although the coil will release smoke if the insulation one winding wire can’t handle to voltage.

      1. The AC line frequency can be very well reached at an ignition coil. Think of a single cylinder two stroke or 2 cylinder 4-stroke engine running at 3000rpm – you get exactly 50Hz. The point is the waveform. You need the flyback effect of cutting the current suddenly or dump a capacitor charged to about 300V suddenly into the primary.
        But the inductance is way to low to connect a sinusoidal mains voltage of 230VAC directly to it.

      2. Interesting points, I did wonder as I was writing the post whether a sine-wave would actually work. The coils with integrated drivers (I believe) take pretty good care of keeping themselves saturated but not burning out etc.

        Anyway, I never said a car coil was the best idea, just that it was an idea… these days there’s probably more of them around than CRT flyback transformers.

  4. If anyone wants to build a serious HV supply let me know, I have a HV transformer that after rectification/filtering will put out 12500v at 1.5 amps. It is three phase in but it can be rewired for single phase if you dont mind ripple.

    oh, it’s 300lbs…

  5. The first reference on the Wikipedia page you listed, Triboelectric effect -> The TriboElectric Series, is very handy if you want to know which of the materials you have on hand will work for you. Polyurethane foam against Teflon being the extreme. Even mismatched materials on rollers can produce a significant electrostatic charge, worth knowing in case you need to avoid doing it too. Last century I designed a wind powered neon light using that effect, but without lowering the costs of the high voltage storage required it was only useful in locations with almost constant wind, and ideally low humidity. The benefit was the low cost and simple construction of the rest of the design, that and it’s longevity.

    1. Nice. Without seeing the detailed spec, I’d guess that one can do everything I’ve ever done. A sample from most fun to still lots of fun would be:
      I don’t know how much protection yours has against high current sparks, but if that’s a possibility then to be safe I’d suggest around 240kilohms, 2 watt or greater rated resistance in series with the output. I usually put it on the ground return side unless your output is floating (no ground), in which case it doesn’t matter.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.