Lightning Detector - tank circuit around 500kHz

Thread Starter

prairiemystic

Joined Jun 5, 2018
307
I'm having troubles with the tuned circuit for a lightning detector. I keep picking up AM radio. Using TA7642 for the amplifier.

I wound 250T on a long AM radio ferrite 1cmx20cm long to get 6.8mH
With no tank capacitor, I pick up at least four AM stations at once, OK for lightning.
1,000pF (61kHz) I pick up three stations one at 580kHz is strong, OK for lightning.
2,000pF (43kHz) I faintly pick up a station and OK for lightning.
10,000pF (19kHz) I don't pick up any lightning or AM radio.
Note this tuning is below the TA7642 150kHz low end.

With AS3935 IC tuned to 500kHz, it also picks up AM radio and false triggers constantly when talk radio happens. Its Inductor MA5532 is 100uH Q=30 DCR=6R.

Looking at other circuits, such as Wenzel uses a filter with 10mH and 1mH inductors.
Another circuit I found tuned to 12kHz or 220kHz and common-base amp (Pedersen circuit)
Silicon Chip/EPE Magazine uses Jaycar LF-102 AM radio ferrite rod with no capacitor - so it's just a pickup inductor.

So I'm puzzled which tuned circuit is best to use. None of them seem to have good selectivity even tuned 10x lower.
 

Attachments

Wendy

Joined Mar 24, 2008
23,415
When I was new to electronics mumble years ago I built a lightning detector that triggered off of static from a AM radio tuned to a dead band(in other words, where there was no station to receive). The louder the static, the closer the lightening.
 

jpanhalt

Joined Jan 18, 2008
11,087
I have used the AS3935 and a home brew circuit. The 3935 has built-in discrimination (a proprietary algorithm) for lightning, and the datasheet gives the look-up table for the intensities. You can modify that table. If you plot the values in that table, it appeared to me that there were two sets of values, and verification of a lightning strike was based on the pattern being within those ranges.

Build a lightning simulator, which basically is a relaxation oscillator, as a start. Then, you can modify it to get a feel for the algorithm.
 

Thread Starter

prairiemystic

Joined Jun 5, 2018
307
I stopped working with the AS3935 because the "discrimination" algorithm does not work here, it picks up male human speech from AM radio and thinks that's an EMI burst. It is triggering based on bursts above a watchdog threshold, with high energy and certain timing. It's weird the datasheet says bandwidth is 33kHz (Q=30) and I get wider than that. So I'm no longer working with the part and rolled back to basics.

Researchers say VHF around 400MHz is also very good particularly for detecting cloud-cloud lightning. But I have no idea how to build an 400MHz AM receiver.
 

jpanhalt

Joined Jan 18, 2008
11,087
500 kHz is an international marine distress frequency. It seems unlikely an ordinary AM station would be on that frequency. What is the call sign for the station you are receiving? What is its radiated power, and how close are you to it?

Since nothing seems to be working for you, there seem to be at least three choices: 1) Start absolutely from scratch (why did you pick 500 kHz); 2) Use the best available chip (AS3935); or Buy a Stormscope (or similar) and see how they work in your environment.

Do you have a question we might be able to answer? In my environment, the AS3935 worked fine, and I had no problem with AM audio at 500 kHz on either my home brew receiver or the AS3935 chip.

Whatever you do, what algorithm do you think is better than the one the AS3935 uses? Do you have records of lightning strikes (cloud to cloud and/or ground) in your area that you can share?
 
Top