Help with settings about kaminari

[shred]

Sure... To change the settings, you invoke the /update endpoint and provide as GET parameters what you want to change. You also need to provide an api_key parameter with the API key you defined in your myWiFi.h file.

For example, to turn off the LED through a browser, you can invoke:

http://kaminari.localdomain/update?statusLed=false&api_key=MySeCrEtApIkEy

(Replace kaminari.localdomain with your Kaminari's host name or IP address, and set the proper API key.)

You can also provide the API key as X-API-Key header, e.g.:

curl -H 'X-API-Key: MySeCrEtApIkEy' 'http://kaminari.localdomain/update?statusLed=false'

[ve3sjk]

Thanks for the response that was the info i was looking for, can i ask if you have seen any real lightning detected using your unit. I have tried 2 different boards in this setup now, in both cases i get perfect tuning at 500k but it does not trigger on lightning events, i have had storms overhead last night and 5 minutes ago with plenty of lightning and the system did not see any of the strikes at all.

[shred]

I'm experimenting with my AS3935 for more than a year now, and I'm finally making a bit of progress. What I could find out so far:

• It's crucial to have enough space between the AS3935 and the ESP board, as the ESP's WiFi antenna is a strong disturber. It seems that 10cm (4 inch) is sufficient.
• A perfect 500kHz tuning seems to be important. After I replaced the wrong capacitors on my CJMCU AS3935, the result seemed to be better.
• My initial idea to auto-adjust the watchdog level turned out to confuse the AS3935 and cause a lot of false positives. I got much better results after setting autoWatchdogMode to false and watchdogThreshold to a fixed value (currently 3 here). I'm about to remove the watchdog auto-adjustment from the code because of that, or at least disable it by default.
• The number of disturbers per minute should be almost 0 on fair weather. If it's higher, give more space between the sensor and the ESP, or find a better place for the sensor with less disturbers. (If there is a thunderstorm approaching, the number of disturbers per minute will still rise, but that seems to be normal.)

With this setup, I only get a few false positives per day. If a thunderstorm approaches, I get many lightning events now. Together with the raising number of disturbers per minute, it's indeed possible to check if a thunderstorm is approaching.

However it's not like that the detector detects every lightning that I can see outside. The distance estimation is also totally useless at the moment, because almost all lightnings are reported as "right overhead".

I keep on experimenting, but it's in the nature of this project that the progress is very slow.