Sketch for Lightning Sensor

[boozz]

@Tore-André-Rosander

You would need a sound detector for the thunder. Time (in milliseconds) between lightning and thunder times 0.34[m].

Same as you would do it with your own sensors (eyes and ears) :-)

BR,

Boozz

[Sparkman]

@Tore-André-Rosander Yes, the sensor can detect the distance. See the details of the sensor here: http://playingwithfusion.com/productview.php?pdid=22&catid=1001. Although I'm starting to believe that it's actually a lightning repeller rather than a lightning sensor as we've had very little lightning in our area after I installed it :laughing:

Cheers
Al

[alexsh1]

I found a similar sensor in the UK and added this to the list of my to be sensors for the weatherstation. Can I use it inside or it has to be placed outside to have a direct vision?

http://www.embeddedadventures.com/as3935_lightning_sensor_module_mod-1016.html

[Tore André Rosander]

@alexsh1 What i read when researching this chip it can be placed inside.

@Sparkman Nice, i see that it has a detection range about 40km so thats already a pretty good distance to get a notification when lightning is detected.

[alexsh1]

@Sparkman I suppose your wiring diagram above is correct? I noted that the sensor is not working with 3.3V. Did you manage to find what the problem was? Looking at the datasheet http://www.embeddedadventures.com/datasheets/MOD-1016_hw_v8_doc_v4.pdf I can see what the module can operate down to 2.4V

I am going to receive the sensor in a few days and test it.

As I have a different module manufacturer, I may need to change the sketch as it uses a different library. This is unmodified example sketch:

*/

// AS3935 MOD-1016 Lightning Sensor Arduino test sketch
// Written originally by Embedded Adventures


#include <Wire.h>
#include <AS3935.h>

#define IRQ_pin 2

volatile bool detected = false;

void setup() {
  Serial.begin(115200);
  while (!Serial) {}
  Serial.println("Welcome to the MOD-1016 (AS3935) Lightning Sensor test sketch!");
  Serial.println("Embedded Adventures (www.embeddedadventures.com)\n");

  Wire.begin();
  mod1016.init(IRQ_pin);
 
  //Tune Caps, Set AFE, Set Noise Floor
  //autoTuneCaps(IRQ_pin);
  
  mod1016.setTuneCaps(7);
  mod1016.setOutdoors();
  mod1016.setNoiseFloor(5);
  
  
  Serial.println("TUNE\tIN/OUT\tNOISEFLOOR");
  Serial.print(mod1016.getTuneCaps(), HEX);
  Serial.print("\t");
  Serial.print(mod1016.getAFE(), BIN);
  Serial.print("\t");
  Serial.println(mod1016.getNoiseFloor(), HEX);
  Serial.print("\n");

  pinMode(IRQ_pin, INPUT);
  attachInterrupt(digitalPinToInterrupt(IRQ_pin), alert, RISING);
  Serial.println("after interrupt");
}

void loop() {
  if (detected) {
    translateIRQ(mod1016.getIRQ());
    detected = false;
  }
}

void alert() {
  detected = true;
}

void translateIRQ(uns8 irq) {
  switch(irq) {
      case 1:
        Serial.println("NOISE DETECTED");
        break;
      case 4:
        Serial.println("DISTURBER DETECTED");
        break;
      case 8: 
        Serial.println("LIGHTNING DETECTED");
        printDistance();
        break;
    }
}

void printDistance() {
  int distance = mod1016.calculateDistance();
  if (distance == -1)
    Serial.println("Lightning out of range");
  else if (distance == 1)
    Serial.println("Distance not in table");
  else if (distance == 0)
    Serial.println("Lightning overhead");
  else {
    Serial.print("Lightning ~");
    Serial.print(distance);
    Serial.println("km away\n");  
  }
}

This is SPI sketch, I'd like to use I2C connection. Generally, the procedure is as follows:

Wait a few milliseconds for the system to stabilise
 Set the tune capacitor to the value indicated on the packaging, by setting the
TUNE_CAP bits of register 8
 Wait 2 milliseconds
 Callibrate RCO by:
o Sending a calibrate RCO direct command (set memory location 0x3d to the
value 0x96)
o Set Register 0x08, bit 5 to 1
o Wait 2 milliseconds
o Set Register 0x08, bit 5 to 0
The factory calibrating tuning cap value will be fine for general use. When you have a
MOD-1016 in an enclosure or close to other electronics it is worth calibrating the tuning
cap again.

[scalz]

can't wait to test this one too ;) i have pcbs for a weather shield (with bme280, veml6070 etc. ) and i have this sensor on it too (https://forum.mysensors.org/uploads/files/1459631385870-stacked.jpg). But not assembled yet, was busy, i hope asap.. i plan to use it with 3v and i2c. @alexsh1 There are libs for i2c if you want :)

Thx @Sparkman for your work on it ;)

[alexsh1]

@scalz yes, please (for i2c lib). What is this red board you have?

[Sparkman]

@alexsh1 @scalz

I'm running my sketch with I2C now, rather than SPI. I found the sensor would lock up occassionally, so recently converted it and am still testing it. I have mine running at 5v now as well instead of 3.3v, but it works fine with either.

I'll post an updated sketch (and wiring diagram) soon. I based both the spi and i2c sketches on the examples found here: http://playingwithfusion.com/productview.php?pdid=22&catid=1001.

Cheers
Al

[alexsh1]

@Sparkman Thanks. Your sketch is very much appreciated. I'll have to adopt it anyway as I have a different manufacture and there is a certain process I have to follow to get it initiated (posted above).

[alexsh1]

Absolutely delighted with the module I have received today by post. The weather is sh@t tonight (lightning, rain etc.) which means that I can test it:

Welcome to the MOD-1016 (AS3935) Lightning Sensor test sketch!
TUNE	IN/OUT	NOISEFLOOR
6	10010	5

after interrupt
DISTURBER DETECTED
DISTURBER DETECTED
DISTURBER DETECTED
DISTURBER DETECTED
DISTURBER DETECTED
LIGHTNING DETECTED
Lightning ~27km away

DISTURBER DETECTED
DISTURBER DETECTED
LIGHTNING DETECTED
Lightning ~17km away

DISTURBER DETECTED
LIGHTNING DETECTED
Lightning ~10km away

I calibrated the sensor and put a reasonable "noise floor". Now I have to convert it into MySensors and hook up to the ceech board outside.

@Sparkman I have my sensor running at 3.3V

[alexsh1]

BTW, I am not sure the sensor is similar to what you guys have.

[Sparkman]

@alexsh1 It looks to be a different breakout board, but for the same chip. So I would expect them to work basically the same.

Cheers
Al

[alexsh1]

@Sparkman it is a different board made (or designed) by a UK company.

We are having a real British summer over here :-)))

LIGHTNING DETECTED
Lightning ~20km away

LIGHTNING DETECTED
Lightning ~20km away

LIGHTNING DETECTED
Lightning ~6km away

LIGHTNING DETECTED
Lightning ~6km away

LIGHTNING DETECTED
Lightning ~6km away

LIGHTNING DETECTED
Lightning ~6km away

LIGHTNING DETECTED
Lightning ~6km away

LIGHTNING DETECTED
Lightning ~5km away

[NeverDie]

@alexsh1
There doesn't appear to be audio input to this board, nor a microphone on the board. So, how does it estimate the distance to the lightening?

[Sparkman]

@NeverDie All these breakout boards are based on this chip: http://ams.com/eng/Products/Wireless-Connectivity/Wireless-Sensor-Connectivity/AS3935. They use a proprietary method to figure out distance and is not based on sound.

Cheers
Al

[NeverDie]

@Sparkman said:

@NeverDie All these breakout boards are based on this chip: http://ams.com/eng/Products/Wireless-Connectivity/Wireless-Sensor-Connectivity/AS3935. They use a proprietary method to figure out distance and is not based on sound.

Cheers
Al

Any impression yet as to how accurate its distance estimates are?

[Sparkman]

@NeverDie No, as we've had very little lightning since I started using it.

[alexsh1]

@NeverDie Al, was quicker replying to your question. :-)
yes, I think it is accurate for my needs - if a lightning detected at, say, 10km+/- 1km this is ok. How accurate do you want it to be?

This is a great technology

[NeverDie]

@alexsh1 said:

@NeverDie Al, was quicker replying to your question. :-)
How accurate do you want it to be?

I'm not sure, but more accurate than lightningmaps.com or it's a no-go. I've played with lightningmaps.com during a few thunderstorms, and although it's great for getting a general idea of lightning strike rate and density, I was surprised that it seems to miss a lot of detections entirely that both my eyes and ears have confirmed. e.g. some strikes that were big enough to rattle the windows didn't show up at all on lightningmaps. Some of the others were detected, but the map plot of the strike was wrong by maybe 10 miles or so.

[alexsh1]

@NeverDie I have not tested this sensor long enough, but recently we have had torrential rain, lightning, etc - you can see some data I published above. The sensor did not miss a single one. There are a few parameters you can alter (noise floor). One has to setup it correctly. I have not started looking into converting the sketch for MySensors yet