I stepped up the build using a 3 Amp power supply, the new side mount i2c board and wirewrap for the Interrupt pin rather than soldering.
https://www.dropbox.com/s/o4pltnlhq91hy … htning.jpg
So far after several months it hasn’t given me any trouble (i2c) buss has been stable.
I also attached a Power/Cooling board, Four channel relay board, and Seven sensor board, (all i2c) to the same Rpi3 and not a bit of communications problems.
The lightning connector demo code for Python and its library work fine for my needs. It dutifully logs events to a log file, which I parse on command with grep when generating Cacti and Mrtg graphs.
I’m going overboard here, but also added a CyberPower UPS with a USB cable and NUT to monitor and control that. Then added an ESP power conditioner with USB line voltage and event monitoring and control that with USB all from the same Rpi3.
I’m hoping to get the red board version of the lightning detector back to compare with the dark gray board. The red board had i2c lockups when using a heavy duty relay board hanging off the USB bus… since now I’m using the dark gray board and a i2c relay board I have not had any problems… so I’m wondering if its the red board version of the lighting detector, or the draw down of the heavy duty relay board that made the i2c bus unstable.
The lightning detection data seems to have been spot on accurate. I have not checked into the ranging data… but I’m not entirely sure how to do that without comparing it to radar time lapses from accuweather or something during the rainy season. Radio being what it is… geography can play havoc with ranging data on non-flat terrain.
I live in Texas and its pretty flat… that should be one set of range data.
But I’m deploying these in the Branson Missouri area… about as mountainous as you can get… so that should be another set of range data.
Its useful to me to just know there is lightning in the area and to cut the mains feed and go on backup power until it passes.