Hello
I have 2 Air modules via 2 different hubs, the 2 modules are about 2 meters apart.
About 40 minutes ago (I wasn’t home but I can confirm with Blitzortnung station) we had a few close impacts detected.
But the 2 modules give other results and count, maybe the fact they are both on my account makes the server go in loops ? Signals would have arrived on the same time on server …
air 4384 detected
and air 7487 detected
This is probably not a server issue. The lightning strikes look like any other data to the server, so if we had a problem with multiple devices on the same network, we’d see a lot more evidence of it.
Rather, I think this is simply the nature of the Franklin lightning sensor itself. That is, it’s very sensitive to position, orientation and the surrounding environment. It should catch most lightning strikes, but it’s not going to catch them all.
If one of those Air units is a field test article, the lightning detector’s antenna might also be mounted in a different polarity than the final shipping Air hardware. I was tinkering with an AS3935 last spring, and you would be surprised how picky it is with its antenna orientation…
Good point but @eric was a late addition to our Field Test crew and he has two production AIR units. Identical hardware.
The Air (AR-00011341) unit that my parents have has had a problem of detecting false positive lightning. I spent quite a bit of time moving it around the property but there was no improvement. In the end, I disabled the lightning notifications.
I currently have it alongside my Air (AR-00005248) for RH calibration. It’s at work and there is a lot of electric equipment nearby so I expect some false positives but my parents Air is detecting way more (930 vs 202). Try did different orientations.
So I’m wondering if Weatherflow can apply a software calibration or de-sensitisation to a lightning sensor?
hello i was going to reply yesterday but was side tracked,anyway i spent over 3 months testing out various locations around our property to find the optimum position of a production air model for lightning. suprislingly the production model mounted horizontal gives me optimum results , this goes against the recommendations of mounting vertically. i also have an old beta model air from early 2017 mounted horizontal both these units are optimized the production air for northerly and beta for southerly . mounting either if these vertical as recommended on the production model produces very poor results unless the storm is on top of you or 5-10km .
for example today we had a storm moving in from the northerly direction it recorded 1900 or more strikes then later in the evening the southerly optimized unit kicked in as the storm shifted south and the northerly optimized unit went quiet .
as i said above i have spent months using these and after much trying of various positions i finally found the sweet spots . these units seem to have a very sharp lobe in directions .
this my capture arrea and currently the northerly optimized horizontally production unit is starting to kick off ! just keep trying and look for the sweet spot…the trade off might be adverse to temperatures .
That’s very interesting data you have produced. I too am convinced that the sensor is directional based on data from two Air devices. I appreciate the time and study you put into this.
How much does compass direction play into distance?
I did disable the production air for now as I got flooded with messages each storm passing by (doubled with the test air … )
Will use the UDP data instead coming weeks to see how it goes as I can do so without a warning on my phone for every x strikes.
What I observed is that the production air is way more sensitive then the pre prod version (sited side by side, same direction) The prod version gave warnings about 10 km before the other one if not more (using also BO to check as you know)
I didn’t play enough to see in how far they are direction sensitive, yet.
Both of my Air devices are, mostly, reporting the same lightning. If they continue, I’ll turn one 90°.
How much does terrain configuration influence on detection capability? I thought my detectors were less sensitive in direction where there are mountains. But could be the orientation, as I read above. Why would orientation change sensitivity? The chip can be configured for higher/lower sensitivity. Maybe WF was testing performance with different sensitivity setting?
Those are questions that have been inquired about before. There have been no definitive answers.
However, it is clear that direction and orientation play a part in detection.
Thanks for the suggestions. Over the past week, the second Air has calmed down but it’s still detecting more. I’ve asked support for help.
We had some lightning in the surrounding area within some showers.
My air didn’t detect anything.
Could a instrumentshelter be an obstruction for the detection?
Before I placed it in this shelter it would even detect lightning more than 35 km away. (I think)
How about turning the Air 90 degreed.
if it is a standard wooden shelter, won’t make a difference, turning it in steps of 90° can improve detection in some direction.
Nothing else that changed in the surroundings that can interfere with detection of the air ? New electric device, even a bit further away that scatters and makes the air deaf ??
Hello Eric.
No noting changed in the surrounding area
Maybe the strikes weren’t that powerfull. 
There are 3 air stations in the surrounding area that also didn’t register anything.
Think CC strikes are not so easily picked up than CG strikes.
And it’s just a wooden shelter indeed.
Will keep an eye on it.
I can always turn it 90 degrees.
Happy birthday !!
Aahh haha Thank you verry much. @spacesnow.
