Shop OBEX P1 Docs P2 Docs Learn Events
Lightning Direction Detector — Parallax Forums

Lightning Direction Detector

william chanwilliam chan Posts: 1,326
edited 2007-11-04 20:05 in General Discussion
Hi all experts,

I need to make a lightning direction detector using 4 plates ( 35cm x 35cm ) to detect lightning direction from 4 quadrants.
To detect which quadrant the lightning is coming from, I need to measure all the plate's voltages simultaneously.
The plate with the highest voltage points to where the lightning is coming from.

Which is the best Op-Amp to amplify the voltage rise from the ions ( could be in a few microvolts to hundreds of microvolts) and which ADC is most suitable for this job?

Thanks.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my

Comments

  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2007-10-17 06:47
    William,

    Messing with Mother Nature’s lightning can be serious business.· Without adequate grounding and electrical precautions you are just asking for trouble·- use EXTREME caution.
    ·


    How far away do you want to detect the lightning? You could probably get away with using 3 plates or sensors to triangulate the signal. If instead of using the plates to detect a signal, ground them and use them as a shield (A metal triangle)... as a sensor use 3 ferrite loop antennas on each face of the metal triangle perpendicular to the metal face, ALL tuned to the same non-local AM radio station.

    As far as reading them all at once, you need multiple processors or a single Propeller. You might be able to use something like an "Anderson Loop" arrangement, but it still would involve multiple processing levels.

    Edit

    Ok, while digging a little further, I found this.... this should get you started.· You can still make the antennas directional by shielding them from on another at 120 Deg


    Basic Stamp Lightning Activity Monitor ... It claims to have a sensitivity of > 50 miles if you mount the antenna 10ft in the air.
    http://timbitson.com/Weather_Projects_files/bslam.pdf

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.

    Post Edited (Beau Schwabe (Parallax)) : 10/17/2007 7:23:46 AM GMT
  • MorrolanMorrolan Posts: 98
    edited 2007-10-17 16:23
    Beau, the link you posted is excellent - I'm now considering building one... smilewinkgrin.gif

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Flying is simple. You just throw yourself at the ground and miss.

    "I think computer viruses should count as life. I think it says something about human nature that the only form of life we have created so far is purely destructive. We've created life in our own image."
    Stephen Hawking
  • JonnyMacJonnyMac Posts: 9,202
    edited 2007-10-17 16:44
    I used to work in the golf industry and with the nation's leading lighting detection company (Global Atmospherics). I can assure you that none of their detectors used your scheme, William. If it works at all, you'll find that it false triggers on a variety of electrical sources.

    Lightning has specific characteristics in the RF and IR wavelengths that distinguish it from other electrical disturbances, and there is difference between cloud-to-ground and cloud-to-cloud lightning. In the golf industry single-point sensors are used to measure the relative distance to a strike (this is used for human safety -- nobody cares about direction, just how far away it is). The National Lightning Detection Network uses a triangular grid of sensors, each with a GPS receiver to keep their internal clocks in sync. When lightning strikes it is typically detect by three or more sensors in the grid; triangulation on relative strength vis-a-vis the sensors determines position.

    In short, lightning detection is not a trivial process.
  • Harrison.Harrison. Posts: 484
    edited 2007-10-17 17:18
    Beau Schwabe (Parallax) said...
    Basic Stamp Lightning Activity Monitor ... It claims to have a sensitivity of > 50 miles if you mount the antenna 10ft in the air.
    http://timbitson.com/Weather_Projects_files/bslam.pdf

    I use that exact design for my personal weather station. It is very sensitive to electrical noise (picks up light switches turning on and such) so you'll need to do some signal filtering.

    Even with false triggers, it gives a pretty good view of local lightning activity. I am able to use the lightning data along with temperature / humidity to detect things like storms before cold fronts. For the curious, my weather station can be viewed at wx.harrisonpham.com/.

    Harrison
    646 x 794 - 97K
  • steve_bsteve_b Posts: 1,563
    edited 2007-10-17 19:44
    We've installed a number of GAI's (now Vaisala GAI) lightning sensors.
    They have 2 types...one that just tells strength of the flash and the other will give direction.
    They are triangulation, as JM says, and they plot maps out.
    The reason they bother with triangulating them at all, is (as far as I know) for insurance purposes mainly. Although forestry has installed a number of these for fire investigating too.

    Here's a link to our lightning map: www.weatheroffice.gc.ca/lightning/index_e.html
    GAI used to have a published map (although it was xminutes old).


    JonnyMac said...
    I used to work in the golf industry and with the nation's leading lighting detection company (Global Atmospherics). I can assure you that none of their detectors used your scheme, William. If it works at all, you'll find that it false triggers on a variety of electrical sources.

    Lightning has specific characteristics in the RF and IR wavelengths that distinguish it from other electrical disturbances, and there is difference between cloud-to-ground and cloud-to-cloud lightning. In the golf industry single-point sensors are used to measure the relative distance to a strike (this is used for human safety -- nobody cares about direction, just how far away it is). The National Lightning Detection Network uses a triangular grid of sensors, each with a GPS receiver to keep their internal clocks in sync. When lightning strikes it is typically detect by three or more sensors in the grid; triangulation on relative strength vis-a-vis the sensors determines position.

    In short, lightning detection is not a trivial process.
    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    <FONT>Steve



    What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
    Lee Travino after the second time being hit by lightning!
  • DaveGDaveG Posts: 84
    edited 2007-10-19 17:03
    William,

    The simplest way to detect the direction of lightning is to purchase a Boltek Lightning Detection System.
    It is fairly expensive but easy to use. You can find it at http://www.boltek.com.

    Several years ago I entered a Lightning Plotting System in the Parallax 2006/2007 SX Design Contest.
    There is a description of it on the Parallax website, http://www.parallax.com/sx/sxcontest06/lightningplot.asp.
    My system was patterned after the GP-1 Lightning Locator developed at Penn State University, http://lcm.met.psu.edu/default.htm.

    The detector consists of two crossed-loop direction-finding antennas, spaced 6 miles apart,
    that triangulate the heading and range of the horizontally polarized magnetic pulse from the lightning stroke.
    A single flat plate antenna is used to determine the polarity of the stroke by detecting the vertical electric field.

    For your application, a simple crossed-loop antenna would give you the direction of the lightning to within +-1.5 degs.
    You could use two AD7821 8 bit ADCs running at 1 MHz to digitize the North-South and East-West loop signals.
    An SX48, with a uM v3 math coprocessor, would be fast enough to calculate the heading, in degrees. There are
    also methods to use simple peak detectors to measure the direction, without any ADCs or microcontrollers, but
    their implementation is fairly complicated.

    Let me know if I can be of any assistance.

    Dave, Apple Valley, CA
  • william chanwilliam chan Posts: 1,326
    edited 2007-10-20 07:02
    Dave,

    Wow, your project looks really complicated !

    In your collected data, how many percent of lightning is electron jumping from cloud to ground, and how many percent is electron jumping from ground to cloud?

    How many volts of amps are you getting on an average lightning from the antennas and what is the spike width ( micro-seconds?)

    I think there should be a way to simplify things a bit....

    Thanks.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    www.fd.com.my
    www.mercedes.com.my
  • DaveGDaveG Posts: 84
    edited 2007-10-20 17:33
    William,

    According to the articles published by Penn State University, 90% of cloud-ground lightning strokes are called
    Negative-Return strokes, where an excess of negative charge (electrons) collects on the bottom of the cloud and
    travels down to the ground. In the other 10%, the electrons move upward toward the top of the cloud.

    A typical cloud-ground stroke produces a vertical voltage pulse. The pulse has a risetime of 2-3 microseconds and
    tapers off to zero in about 200 microseconds. It travels at the speed of light in all directions. You can detect this vertical
    voltage pulse by using two flat plates mounted horizontally to the ground, one above the other. The bottom plate is
    grounded to the earth and you measure the voltage between the bottom and top plate.

    For example, if the plates are 18" x 18" square, and 18" apart, a stroke that is 60 miles away will produce about 3 volts
    between the plates. The voltmeter must have a very high impedance so it doesn't short out the plates. Plus it must have
    a fast response time to catch the 2-3 microseconds risetime of the pulse.

    There are many analog and digital circuits that can capture and display this pulse. A high speed digital storage oscilloscope
    will also do the job.

    Horizontally mounted plates cannot be used to sense direction of cloud-ground strokes because you will get the same reading
    regardless of the direction of the lightning. Turning the plates on edge does not do the trick either because the voltage pulse
    is vertical, and only plates that are mounted horizontally will detect it.

    Luckily, physics comes to the rescue. A cloud-ground stroke has two components, a vertical voltage pulse and a horizontal
    magnetic pulse. The magnetic pulse can be sensed by a pair of crossed-loop antennas to compute direction.


    Dave, Apple Valley, CA
  • william chanwilliam chan Posts: 1,326
    edited 2007-10-22 05:10
    Dear Dave,

    Is your lightning detector still in working condition after so many years?
    In your case, are you sampling 1 mega samples per second?
    Can your board and ADC detect the Ground-Cloud stroke?
    Does a Cloud-Ground stroke produce a positive voltage on the upper plate and a Ground-Cloud stroke produce a negative voltage on the upper plate relative to the lower plate?
    Should I bias the plates to 2.5v steady state so that it can detect both strokes?

    I think a 10Mega ohm resistor between the 2 horizontal plates may be required to zero any floating voltages, am I correct?
    Why can't we use the physical ground ( the soil on the ground ) as the lower plate?

    Thank a lot.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    www.fd.com.my
    www.mercedes.com.my
  • DaveGDaveG Posts: 84
    edited 2007-10-23 23:32
    William,

    I sent you a Private Message on the forum yesterday. Let me know what you want to do.

    Dave
  • pwillardpwillard Posts: 321
    edited 2007-11-04 20:05
    Has anyone located the BSLAM source code for the Stamp? It now seems missing from Tim's Site. (embedded link to get the code in the PDF... ISN'T)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

    There's nothing like a new idea and a warm soldering iron.
Sign In or Register to comment.