Lightning Detector

Discussion in 'Embedded Systems and Microcontrollers' started by BR-549, Jul 18, 2016.

  1. BR-549

    Thread Starter Well-Known Member

    Sep 22, 2013
    1,981
    388
    jpanhalt likes this.
  2. Willen

    Member

    Nov 13, 2015
    138
    12
    Just a tiny chip (without proper antenna) detects lightning and rejects other similar interference signals, sounds awesome! I love such 'Bio-engineering' :)
     
  3. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,127
    266
    Cool- Ok, so what do you do with this thing?
     
  4. BR-549

    Thread Starter Well-Known Member

    Sep 22, 2013
    1,981
    388
  5. #12

    Expert

    Nov 30, 2010
    16,248
    6,745
    I tried this once. Connect to power line with a 100k resistor and 1000 pf to ground. Repeat once, look at the results on the 'scope.
    Plenty good enough information to shut off machinery when storms are 5 miles to 10 miles away.
    Of course, that was my goal, machinery protection. If you want to know all about where the lightening is, which polarity, and what color, this chip does a lot more than the experiment I did. Just, it isn't necessary to use a sophisticated chip to detect one of the most violent energy discharges on the planet.
     
    BR-549 likes this.
  6. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,672
    899
    Definitely an interesting chip that will be included in my next DigiKey order. Many years ago, Ryan developed what he called a Stormscope for locating storms based on lightning strikes (aka Ryan Stormscope). Its early adoption was for general aviation and single-engine aircraft, because it did not require a radar dome. It then became an accessory even in aircraft equipped with radar, as there was evidence that turbulence correlated better with lightning strikes than with rainfall -- radar detects rainfall. Ryan's company has been bought several times since then. Modern advances in avionics and tracking may have made the Stormscope somewhat obsolete (based on my brief review of comments to that effect).

    Here are some general links you might find interesting:
    https://en.wikipedia.org/wiki/Lightning_detection
    https://www.stormscopes.com/stormscopes.html
    http://www.bennettavionics.com/wx10.html

    As for detecting storm fronts vs. lightning strikes, I would quibble with that distinction.

    Because radar detects rain, one might also say that radar detects the leading edge of a rain storm. It is well known with low-power radar that a front of heavy rain will attenuate the signal so much that a following storm of even heavier rain may not be detected. After a brief review of the datasheet, it appears that a similar phenomena may happen with this lightning detector (and others as well). Lightning on the advancing edge of the storm is detected and ranged. Hence, the edge of the storm is said to be "detected." Signals generated within the storm may not be as well resolved with respect to distance.

    Anyway, I now have a fun project for the Winter. Thanks.

    John
     
    BR-549 likes this.
  7. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,672
    899
    My order for the chip and a QFN adapter has been placed. Incidentally I added a PIC 24F16KA102 because it was cheap. Next update after the first snow in Cleveland.

    John
     
  8. upand_at_them

    Active Member

    May 15, 2010
    246
    29
  9. BR-549

    Thread Starter Well-Known Member

    Sep 22, 2013
    1,981
    388
    #12, what kinds of signals did you get on your power line setup?
     
  10. crutschow

    Expert

    Mar 14, 2008
    12,977
    3,220
    Yes.
    I'm also interestested in the waveshape and amplitude of the signal. :)
     
  11. BR-549

    Thread Starter Well-Known Member

    Sep 22, 2013
    1,981
    388
    I'm wondering if he monitored for HF bundles coming down the line, or the level of noise floor.
    Or both perhaps.
     
  12. #12

    Expert

    Nov 30, 2010
    16,248
    6,745
    I got short, positive going pips. Maybe a volt in amplitude, maybe a few tenths of a volt, but I really don't remember because I did that in 1980. The amplitude was easily in the range of an op-amp input. I knew I could amplify the pips, so that's all I needed to know.

    The principle is that I'm using the entire local power grid for the antenna. There was no attempt to locate or evaluate the quality of the lightening. The criteria was: If x number of signals, like a dozen or so, happens in one minute, it must be lightening. Florida thunder-boomers have been clocked as high as 2 strikes per second. I could detect pips 5 to 10 minutes before a storm arrived, so a one minute integral was sufficient to make a decision.

    The RC time constant of the filter was 100 microseconds, so you have to know that all I got was the high frequency components...about 10KHz and up.

    ps, the noise level was undetectable on a scope set for 1V per division.

    Crap. Now I'm beginning to doubt myself. Was that a high pass filter? I believe so, therefore it must have been capacitor input and resistor to ground, twice. I had a drawer full of 1KV 1000 pf ceramic caps, so that's what I used.
     
    Last edited: Aug 17, 2016
    BR-549 likes this.
  13. BR-549

    Thread Starter Well-Known Member

    Sep 22, 2013
    1,981
    388
  14. Rabbit H

    Member

    Aug 5, 2016
    88
    10
    If all you want to do is detect an approaching lightning storm, tune an AM* radio between stations and feed its audio output to an alarm generator via a threshold filter (squelch circuit). This will tell you when lightning is within range. If you want direction also, pick up an old aircraft ADF system (receiver, indicator, and antenna) on Ebay. They're rapidly becoming obsolete for aviation nowadays and can often be had for a song.

    *For those of you who don't know what an AM radio is, ask your grandparents to tell you how they used to listen to baseball games. :D

    Harvey
     
  15. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,672
    899
    Just received my AS3935 chip from Mouser. I didn't notice there was a center pad until shortly after I ordered it. Here is a clue, a 0.65 mm pitch QFN/QPN adapter board will work fine for the peripheral pins, but the pads extend into the area of the solderable center pad. BTW, Mouser lists the price as "each" for the adapters, but I got two adapter boards that had not been separated for $6.00.

    Here's a valuable ($13.65) clue: get the ThunderClick board. It comes with the chip and a small coil antenna already attached.
    Mouser # 932-MIKROE-1444 @ $35. When I noticed the center pad, I figured I could grind the pads on the QFN adapter to be a bit shorter, and I may still do that, but the price difference is not enough to make that choice with open eyes.

    John
     
    BR-549 likes this.
  16. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    You can do a minimalist job with a MK484 TRF radio tuned to 300kHz - I think the current part is the Toshiba TA7642.
     
  17. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    I think it was an NS applications engineer published advice how to deal with those DIY style.

    The base pad usually implies some form of heatsinking - you have as much copper as possible on the non-component side, and a hole for a stub of the thickest tinned copper wire you can lay hands on. Heat the stub with the iron while its pressed onto the pad, and run some solder in. Finish up by making a regular solder joint between the wire stub and the copper area.
     
  18. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Apparently; 300kHz is a prominent frequency component in lightning strikes.

    An overhead storm usually hits something in the back gardens of the houses over the road - once or twice, the nearby strikes have damaged computer equipment in my flat.

    A regular AM/FM portable is adequate for a warning device, but MW & LW pick up too much crap near the PC - usually I can find a quiet spot on one of the SW bands that can let me hear the crackles.
     
  19. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,672
    899
    The adapter pad ("Schmartboard") has a PTH in the center for that wire nub. All one would need to do is grind down the pads about 0.1 mm each to whatever suits you to clear the non-leaded connections on the underside of the chip.

    BTW, the adapter board is grooved and pre-coated with solder for the device pins, so positioning is really easy, and I suspect soldering would be too.

    John
     
Loading...