Weather-Watch.com

Weather Equipment/Software => Home Built/Kit Weather Stations/Sensors => Topic started by: MichaelPT on May 03, 2017, 12:44:09 AM

Title: Franklin Lighting Sensor AS3935
Post by: MichaelPT on May 03, 2017, 12:44:09 AM
Hi

As any body tried  this  project out?

www.arduino.org/learning/tutorials/boards-tutorials/franklin-lightning-sensor-as3935

I have af ranklin-lightning-sensor-as3935
Will a arduino uno work ?

website info they used a
Arduino Yun board

Mick
 
Title: Re: Franklin Lighting Sensor AS3935
Post by: teal on May 03, 2017, 09:22:46 AM
  Not sure if any of these can be used in WD    http://www.techlib.com/electronics/lightning.html  I have made three of these, two  can be seen there  https://flic.kr/p/s45YVT  &  https://flic.kr/p/sZYurD  , but I'm not sure how to connect them to WD or if it is possible to do so, they have a range of about 30/35 miles. If any knows how to connect them to WD (if they can) that would be good.

  teal.
Title: Re: Franklin Lighting Sensor AS3935
Post by: niko on May 03, 2017, 01:48:22 PM
  Not sure if any of these can be used in WD    http://www.techlib.com/electronics/lightning.html  I have made three of these, two  can be seen there  https://flic.kr/p/s45YVT  &  https://flic.kr/p/sZYurD  , but I'm not sure how to connect them to WD or if it is possible to do so, they have a range of about 30/35 miles. If any knows how to connect them to WD (if they can) that would be good.

  teal.

Those circuits are essentially the same function as the front end of the Hobby Boards lightning sensor. If someone has the original Hobby Boards schematic it would not be difficult to adapt it to use one of Charles Wenzel's detector designs as the front end. 
Title: Re: Franklin Lighting Sensor AS3935
Post by: teal on May 03, 2017, 03:31:28 PM
  Hi Niko, Right I see, I will have to have a look on the internet to see if I can find any thing about it.

many thanks.

  teal.
Title: Re: Franklin Lighting Sensor AS3935
Post by: niko on May 03, 2017, 11:59:09 PM
The front end (hopefully) generates a pulse for each strike, the rest of the Hobby Boards unit is a 1-wire counter that counts the pulses and provides the interface to the 1-wire bus so the count can be read by e.g. WD. I think a schematic may be in the Bitson Weather Toys book, but IIRC the final Hobby Board was not quite the same. ***

Be aware that these (Wenzel/Hobby Boards) are simple broadband receivers with very limited filtering so almost anything that generates an electrical discharge will set them off. Electric fences, welders, and heavy duty relays/contactors switching loads like motors can result in false readings. You might have something in your garage that sets it off  :wink:

The AS3935 chip which started this thread is a more sophisticated animal. I believe it's used in, or can be added to, the latest version of the blitzortung lightning device, and is used in some portable/pocket lightning alert devices. I was excited when it was introduced but have no personal experience with it. From what I have read it needs some tinkering with the settings to get a good result.

*** Update: Found this (http://www.ishikawa-lab.com/pdf/Lightning.pdf) after posting
Title: Re: Franklin Lighting Sensor AS3935
Post by: mihec on May 11, 2018, 06:39:48 AM
I'd like to reopen this thread for further discussion. I'm using this sensor in the 'PlayingwithFusion' package since last year. Results I get are great for the price of the package.
I have two sensors connected to:

For both I'm using the publicly available code which differs in a small detail that calculates the energy. The set with a RPi works reliably with almost no false lightning detections. I was not really believing the distance that the sensor reports so I just use the energy of each strike for plotting. The output is reasonable and shows how the storm approaches and leaves. One can also tell if the storm was just passing nearby.
The set with an Arduino board is connected to the same RPi which provides the power to Arduino and logs the lightning from Arduino. The code I use there provides the distance and the energy. Sometimes both values are a bit confusing. My question is if the distance and the energy values can indeed be trusted to some extent or is it better to rely just on the energy that the sensor reports? Same sensor is used in the Weatherflow station and they output the distance only. My feeling is the distance is not to be trusted (but maybe they have a more sophisticated algorithm implemented).
I tried to link the sensor with WD. It would work but by writing a "bridge" from simple 1-value output to a Lightning2000 compatible format.
Title: Re: Franklin Lighting Sensor AS3935
Post by: mihec on September 01, 2018, 08:19:59 PM
For completeness, I want to add more information about this project.

I've made a progress in adapting the output of this sensor to a Lightning2000 file format. The file is uploaded to WD's directory every minute. WD does register the file but the total number of strokes does not match with my raw file (sometimes a factor, not an offset).

My third detector is soldered with an Arduino Nano and connected directly to the laptop where the WD runs. I've made a simplistic Python script that listens to the USB port with Arduino and writes the data to a log file + a file with the timestamp of the last stroke. I use one of the customclientraw's fields (e.g. the forecast) and load the timestamp's file content. This way, I can display almost real-time stroke detection.

It would be great if WD had an option to read strokes from Arduino directly to avoid Python middle layer.