One of the driving principles of a lot of the projects we see is simplicity. Whether that’s a specific design goal or a result of having limited parts to work with, it often results in projects that are innovative solutions to problems. As far as simplicity goes, however, the latest project from [153armstrong] takes the cake. The build is able to detect lightning using a single piece of equipment that is almost guaranteed to be within a few feet of anyone reading this article.
The part in question is a simple, unmodified headphone jack. Since lightning is so powerful and produces radio waves in many detectable ranges, it doesn’t take much to detecting a strike within a few kilometers. Besides the headphone jack, a computer with an audio recording program is also required to gather data. (Audio is often used as a stand-in for storing other types of data; in this case, RF information.) [153armstrong] uses a gas torch igniter as a stand-in for a lightning strike, but the RF generated is similar enough to test this proof-of-concept. The video of their tests is after the break.
Audacity is a great tool for processing audio, or for that matter any other data that you happen to be gathering using a sound card. It’s open source and fairly powerful. As far as lightning goes, however, it’s possible to dive far down the rabbit hole. Detecting lightning is one thing, but locating it requires a larger number of weather stations.
hmm, i think i will try this in the coming rainy season.
Just tune to a dead area on the AM band of a radio and you can listen to the strikes from miles and miles. Hook that to a computer if you want.
Check out I believe Episode 1 of ” Wildest Weather with Richard Hammond ” .. He actually gives a live demonstration of this.. Really cool!
You mean trilateration. Very common mix-up.
If you’re interested in lightning detection check project Blitzortung, it has pretty cool PDF with explanation how system works. It’s a challenge to get reliable detection (in fact impossible with single station), that’s why BO uses many stations that send data to the hub where it is compared and false positives are rejected. Also positioning is done there, evey station has GPS receiver for precise timebase so that lighning positions can be determined depending on how much time it took for radio wave to travel to different stations. Project is pretty amazing and precise.
I have an easy lightning detector: most DVB-T recievers.
Try watching television in the thunder, and the picture just freezes. The higher the compression ratio, the longer it takes to resume again after an unrecoverable error, so in practice OTA television becomes completely useless thanks to being digital – the same reason why DAB radio sucks donkey balls. The effect can be reproduced with piezo lighters, light switches, someone calling the elevator, frayed spark plug wires in a moped going down the street…
For close lightning storms it may be possible detect the rf signal and on the other channel record the audio to capture the thunder clap than determine the time difference between the rf and audio apply the appropriate conversion factor for your preferred measurement system and you’ve got the distance determined
How do you know the sound is from that particular source and not some other strike that happened even further away, that is just reaching your position?
Yes that is an issue that needs solving ;)
There may be a correlation in the two signal amplitudes that allows you to pair them. You can also have a ring of LEDs and pickup the direction of the light flash too. More than one ring with sufficient spacing between them would have an even more robust direction finding accuracy. Keeping your station clocks synchronised as accurately as possible is important if you are going to use sensor arrays of any form.
This is a very cool idea. Whoever builds a dual-mode lightning/thunder detector/ranger, let us know. Guaranteed Hackaday writeup. :)
The older NDB/ADF (Non Directional Bearing/Automatic Direction Finders) indicators in aircraft have often been used to find the direction of lightning strikes – low end of the band on the NDB is best when you’re not listening to the ball game (these work in the AM band so you can get commercial radio on them). There is also a commercial version of this (“Stormscope”) that provides a radar-like display of lightning, though in the age of real-time doppler data it’s less popular than it used to be.
The Austria Microsystems AS3935 uses a 500KHz resonator-as-antenna to detect strikes w/ RF. I have a module I bought on tindie and it works, although my luck has been sparse lately (think my new battery boosterpack has too noisy a buck or something).
I haven’t tried the already built module, but i received samples of that chip from Austria Microsystems, and also samples of coilcraft of the coil mentioned in the AS reference design. Lastly not much spare time to sew things together
The tricky part for me was writing a microcontroller procedure to auto-tune the gain control settings and play with various strategies. Of course once you change it, you have to wait for the next thunderstorm to test. What tipped me off about my power supply issue was how the system I wrote kept leaning towards maximum squelch and it’d still fire off false alarms on a clear day… whereas the prior year before my 1st gen LiPo TI launchpad boosterpack stopped working (think the battery sat too long fully drained) it was a better behaved system.
One of these days I’ll get that back up & working…
UPDATE:
Hello, this is 153armstrong and we have a quick update. We tested the headphone jack with an actual Spark Gap and have just posted the results. You check out here: goo.gl/wCXD9W
https://thehackerdiary.wordpress.com/2017/06/10/the-headphone-jack-meets-an-actual-spark-gap/
Cool project! Thanks for the update.
I recall an Amateur Scientist column from the 60’s that feature 2 inexpensive AM transistor radios oriented with the ferrite rod antennas oriented at 90 degrees from each other and the headphone output connected to a coincidence detector. This allowed detection direction finding and detection of strikes several hundred mile distant by sky-wave propagation.
Check the Scientific American book of projects for the amateur scientist, with material from 1950 to 1960 https://www.amazon.com/Scientific-American-projects-amateur-scientist/dp/B0006AWNUE
The section on weather stations: They call this “sferics” as in atmospherics. The author claims that the ratio of counts between three frequencies – like 430 kHz, 2000 kHz, and 5100 kHz, gives a measure of distance to a storm.
Anyone who reads HaD should have a copy of this book, and there are PDF’s out now with the file name projects_for_the_amateur_scientist.pdf You will be amazed by how many HaD topics are covered very well in this 60 year old material. A great project will be to update the whole thing with current tech and methods. Who is up for it?
If you are looking for a nice project with an AS3935 and an Arduino nano, take a look at this site:
https://www.pi4raz.nl/razzies/razzies201701.pdf
They describe a very nice lightning detector with a display and several possibilities to change different parameters.
It is in Dutch and English.
I know it’s silly, but I read this bit:
‘The build is able to detect lightning using a single piece of equipment that is almost guaranteed to be within a few feet of anyone reading this article.’
And instantly my brain went ‘Ears!’.