Expert: Daniel Mazur - 1/17/2010

Question
When I run a high voltage current through a single ordinary wire, there is no obvious interference heard on a nearby unrelated radio receiver. But if I run the exact same voltage and current and make it jump through the air via a break in the wire, i.e. a spark gap, a very loud and audible interference is heard on the radio set. Why does this happen? Why does the unbroken wire suppress the electromagnetic interference, whereas the spark gap does not? Don't all moving charges (electrons) generate an electromagnetic wave, regardless of what medium they are travelling through? Thanks!

Answer
Hi Eddie,

let me start with this: Interference happens in both cases, the difference is in the magnitude of the effect. Secondly, there are two kinds of interference happening at once in both cases, the _static interference_, which is proportional to the magnitude of the magnetic field induced around the test wire (broken or not), and _dynamic interference_, which is proportional to the RATE OF CHANGE of the induced magnetic field.

The static interference is generally small. This is because a static magnetic field is NOT a wave. Static field by definition is "just there". It influences CRT screens, wires and semiconductor elements (through magnetoresistance, Hall effect), but it decreases as R^(-4), where R is distance from the current-carrying loop! Whatever realistic current you may be passing through an unbroken wire (yes, be it several hundreds of amperes as in the biggest electromagnet in existence), the effect a few steps away will be minimal. The current flowing through a broken wire BEFORE the spark jumps is subject to the same considerations. So, the _static interference_ is negligible in both cases.

If your high-voltage high-current power supply has a switch that is protected against sparking, then switching the PS on will cause the sparking interference with instruments next to it as well. Most DC power supplies have a capacitor connected after the switch, which is guaranteed not to spark up to and beyond the maximal voltage of the PS. Then the PS itself won't interfere with other instruments and we can discuss the continuous and broken wire alone.

Now what about the dynamic interference? We have a PS that assures that the current supplied to the wire doesn't rise too fast (time derivative of the current is lowered) and in the first fraction of a second a normal current flows in the broken and unbroken loop alike. The unbroken loop current rises to I=U/R withing some hundredths or tenths of a second and stays that way until we interrupt the circuit. The broken wire current will first rise (thanks to the PS discussed above) and then drop slowly to zero as the halves of the broken wire charge. The dynamic interference in both cases until this point is limited by the maximal time derivative of the current(!) and will be limited by the PS and the speed of charging of the broken wire. The interference of this can be observed by sensitive instruments, but in most cases the other instruments won't notice it, as it gets lost in the noise measured anyway.

Now, what happens to the broken wire: After it charges up so that the electric field (voltage divided by the distance of the bare ends of the wire) exceeds the ionizing field of air. In this field the air suddenly (microseconds or even faster) ionizes, by which it becomes as conducting as any conductor. Once this happens, all the charges accumulated on both ends of the gap flow to the other end across the gap. The peak current thus created is large, yet more importantly, it is a current PULSE, so the current's time derivative is huge. So the induced magnetic field also has a large time derivative and THAT causes induced secondary currents in the nearby electronic devices.

Tu sum up, I repeat that it's not the magnetic field itself that is picked up by electronics as interference, it is the time derivative of the magnetic field. The time derivative is much larger in a spark than in any DC loop. I haven't used any equations in this explanation, I don't think they were necessary. If there is more unclear about it, please post a follow-up question.

All Best,
Daniel