Electronics -> Lightning Detection

This page has had views

About Lightning Detection

Before we get into this topic, take a look at this video, from YouTube. The poor guy in the vid is working away on his motorbike when lighting hits a building across the yard from where he's working. His reaction is understandable! If you watch it carefully I'm sure you can see a spark jump from the bike to his hand.

Ever since being a kid, I was fascinated by lightning- one of the most powerful and destructive forces in nature. This led to development over the years of a number of detectors, and this web page is about these.

The object is to detect when storms are approaching, or that there is a lot of activity around. Primarily there are two methods of detecting nearby lightning: by electrometer, and by RF emission.

Electrometer Method

An electrometer is essentially a meter for measuring electrostatic charges. There is a constant electric field in the air which rises from zero at ground level and goes upwards- in fact we live in a soup of electric charges, as you will see if you experiment with this little circuit. Any construction technique will do, but the gate lead of the FET should be extremely well insulated; you can use the teflon insulator here from an old N-type RF connector. The antenna consists of a small piece of copperlcad board about 3in square, mounted up in the air on about 6in of stiff wire, with the copper connected to the FET gate via the 10 meg resistor. The circuit should be grounded for best results. The FET should ideally be an Insulated Gate device, however JFETS of the 2N3819/BF244 variety seem to work. The meter ideally is a centre-zero instrument if you can get hold of one, if not take an ordinary 100uA meter movement and see if you can move the needle to the centre using the zero screw; on some there is a similar arrangement at the rear of the movement for anchoring the rear spring that may allow more range of control.

To check it out, set the sensitivity control to maximum resistance, and switch on. Set the meter to center zero with the set zero control. It will probably drift a bit after initial switch on. Now get a plastic rod such as bic pen and rub it in your sleeve a couple of times to charge it. You will see that the meter responds with deflections first one way and then the other, as you move the rod toward or away. Increase the sensitivity as necessary, try just walking around near the meter... If you take it out of doors, notice how the meter reading changes as clouds go by. If a storm brews, you will see a huge build up of charge, and violent movement of the needle- frequently from end-stop to end-stop- when lightning occurs. Just a couple of things to note about the (very basic) circuit above; 1. there is nothing stopping you setting the sensitivity pot to zero resistance and maybe damaging the meter movement. You could put a 1K resistor in series which would both protect the meter and set the maximum to something sensible. 2. The FET is protected only by the 10M resistor and will die if you put a static charge from any seriously charged up object on it.

The trace below is a simulation of what you see if you feed the output to a plotter (which is possible) as a storm passes near. The sudden spikes are excursions in charge caused by lightning, and the huge one near the centre of the trace is what happens when the lightning is quite near.

There are better ways to do this, and circuits have been published on the web using special op-amps. Best of all is a device called a field-mill- try Googling this. However, fascinating though the electrometer is, for a small portable device the RF method is more appropriate.

RF Method

You will be aware of how a domestic Medium Wave (BC Band) radio crackles when storms are near. This is due to RF radiation from the lightning, which occurs during the actual lightning stroke when rapidly varying currents of several kilo-amps (kA) sweep along the lightning channel. The radiation is predominantly at very low frequencies (VLF) and is commonly termed 'sferics', an expression which comes from the earliest days of radio and is a shortening of 'atmospherics'.

The best way to detect them is to tune a radio receiver to an empty frequency somewhere in the LF region, and for small portable devices a frequency around 40 kHz is fine. I first built a device of this type in the late seventies, and it became called 'Lightec' for lightning detector. This had purely a visual display, an LED that got triggered on for a second whenever a sferic was received. The success of this device led to the development in 1984 of Lightec AV mk2 shown in the photos below. It operates on 40 kHz, and has both an LED and a speaker so you can listen to the sferics- a strangely relaxing exercise. You can select either mode, hence AV for audio-visual. The unit is powered by rechargeable cells and is built in the case of a pager, using the pager's audio transducer as a speaker.

This little box is 24 years old now and still going strong; and it had an overhaul in 1994 and new cells fitted- but you can't get these 90 mAH DEAC cells any more so I'm considering converting it to use lithium coin cells instead. But a more up to date project is underway as well, to provide a microprocessor controlled device that will provide lots more information.

In the late 1980's I built a modified version of the Lightec where instead of flashing an LED, the monostable circuit was modified to pulse an electromechanical counter. Parts of this unit still exist, however trials came to an abrubt end when, during battery charging using an accidentally way too high current, one of the battery cells exploded blowing a hole in the side of the case and largely destroying the electronics. I'm just glad I wasn't around at the time. However, during it's life, some really interesting results were obtained. The unit was sited in an outbuilding to cut down on false counts from thermostats etc and I used to read the counter each day, subtract the previous day's count and plot the result on a graph.

I started work on replacement a few years ago, and the idea was to log not just the total count but also to provide a history so that the user could see whether activity was increasing or decreasing. The prototype uses a three-stage radio receiver tuned to 40 kHz (no surprise there then) and this feeds into a PIC16F84 microcontroller, which does the counting.

I must emphasis that this is a prototype device, and the display isn't so user friendly but in the picture above, the top line is showing the total flashes recorded, and the bottom line is showing the history. This is a row of digits, each representing a minute ago in time. Up to 9 counts in a minute can be counted, and after one minute this is passed to the digit to the right, then to the one to the right of that a minute later etc so a history over the past ten minutes is displayed. In the photo, it's showing a total of 6 counts, with 1 three minutes ago, 1 five minutes ago, 2 seven minutes ago, and 1 ten minutes ago. A further count has dissapeared off to the right. The complete system is shown below. The consumption is about 5mA at 9v so the MN1604 battery shown (manufacturer quotes 550 mA/H) should run the unit for 110 hours which is plenty enough for casual use.

I hope you enjoyed reading this, if you've any comments or questions I'd be glad to hear- just email me at alancordwell[at]blueyonder[dot]co[dot]uk- with the things in square brackets replaced by dots and ats etc...

Last Modified 10/8/2008