lightning detector

This is a lightning detector circuit for my weather station.

The properties of lightning have been well documented.

The EMF spectrum from lightning peaks at about 5 kHz
and drops to low levels above 10 MHz.
Some detectors focus on one frequency say 300 kHz:

Others detect DC to a high frequency.
I decided to have an amplifier from about l0 Hz to 100 kHz.
I added a monostable of 0.5 seconds to stop multiple detections
from one flash.

To interface to the arduino I use interrupts
Lightning can be detected at any time so this seemed the best option..

// simple program to show lightning sensor interface on arduino 2009
// int lightning will be updated on each lightning detection
//pin 3 causes an interupt and lightning_interupt runs
//at any time in the main program

volatile int lightning = 0;
const int interup_pin = 1; //this uses D3 as the interupt

void setup()
Serial.begin(9600); // initialize serial communication with computer

void loop()
delay(2000); //this is the main program working away
attachInterrupt(interup_pin, lightning_interupt, FALLING); //allow interupts
Serial.println(lightning); //main - here all we do is print the int

void lightning_interupt() //this program runs on interupt on pin D3
detachInterrupt(interup_pin); //disable interupts only need one(may bounce)
lightning +=1; //increment the int lightning

I wonder if one opto-isolator is enough for safety!

One half decent lightning strike and that whole lot is carbon :o

It's not really designed to prevent a full strike from reaching the house... but it does have a more direct path to earth ground than leaping through the opto. Most real world lightning protection for underground wiring is designed with sacrificial parts.

This is a nicely modified version of what Tim Bitson and others use to count strikes.

Nothing stands up to a direct lightning strike.

If that happens kiss every thing goodbye.

Mainly we are trying to keep large static voltages from our arduino.

This does not even need lightning. If you have wire antenna
you can get large static charges from clouds and even on
clear days a decent spark can be generated on ocassion.

pwillard - thanks

yes I do have Tim Bitson's book. It' been very usefull.

I tried his lightning circuit and it works fine.
Mine is more sensitive and stops multiple counts for single

Plus of course I'm not using 1 wire.

I designed a circuit board for this detector.

The eagle and gerber files are here:

download from

hedge your losses. use optic send / recieve to your arduino. that way if it gets nuked you just loose anything optic side of the antenna.

the same sort of rig is used in musically tuned tesla coils. keyboards are connected to the tesla coil via an optic line.

You are so right.

Just 2 days ago we had a massive ground strike not 200 m from
my house.
The strike was on a large 30 m high tree.
The explosion was spectular. Nothing was left of the tree except a stump.
We found parts of the tree some 100 m away.

It came out of the blue without any preceeding rumbles.

I lost my old desktop whick was connected at the time.
A good excuse to buy a new laptop.
Nextdoor lost a modem.

The arduino was connected at the time to the remote weather station but I was lucky there.

This is the first near miss after 15 years here so I hope it won't happen again for another 15 years.

If you have access to silicon chip magazine (Australia).

This circuit is described on page 61 of the December 2010 issue.

what resistor watt ratings did you use?
what about capacitor voltage ratings?
or can one use run of the mill parts and not worry about it?

thanks Pete

There are no special requirements for the components.

The static induced voltages have very little power.
The input resistor R6 - 220 k drains the static voltage and a 1/4 watt is fine.

If you win the 1 in 1,000,000 lottery and get a direct strike you can write off
the $5 investment(tax loss ?).

thanks - I've ordered my $5 in parts. I hope I don't win the direct strike lottery...