Receiver performance drops by 90% at low temperature

[lck]

Hi,
I have been running a receiver on the roof since a few months. The outoor antenna (Vinnant) is connected to a Nooelec NESDR SMArt v5 which is connected to a Raspberry Pi 3B+. Both devices are on the roof as well in a weatherproof box. Cables exit at the bottom of the box.

Today evening I noticed a very low count of ACs received. I started to dig. Find attached stats-screenshots as well as the temperature curve (blue line). It seems that as soon as the temperature drops below approximately 7°C the messages-received-counter drops dramatically. You can see that today during daytime numbers were normal again. Only close aircraft are seen and their messages arrive with the usual strong signal strength.

The Raspberry Pi reports a CPU-temperature of 30°C and the SDR USB stick is usually very warm when active, so I doubt it is too cold for the two devices.
But what else could cause this?

Has anyone experienced a similar issue before?

Thanks
lck

[lck]

Today morning (3°C) range and AC count look normal again. So if it isn't a temperature issue, what else could it be? humidity on the antenna?
Thanks

[wiedehopf]

Inspect inside the box for humidity?

Put some silicone grease / dielectric grease into the antenna connector.

Really hard to tell anything.

Running a standalone decoder + graphs will give you better statistics than the FA statistics.
Might give more clues but might not.

https://github.com/wiedehopf/adsb-scripts/wiki/Automatic-installation-for-readsb

https://github.com/wiedehopf/graphs1090#graphs1090

[nikall]

Check antenna cables. If you don't use connectors in every part of your cabling then do it. Soldering or knitting coaxial is a very bad idea (that's from personal experience).

[lck]

Today i finally managed to get on the roof. I dis- and reconnected the antenna cable on both sides (no solder, standard connectors). No visible traces of humidity at all - neither in the box nor on the connectors. However, after doing so the count of ACs went back to normal. So it must have been a loose contact issue. Still strange since both connectors of the antenna cable were fastened normally. If the issue pops up again I am going to follow wiedehopf's advice and put some grease into the antenna connector.

[lck]

Here is a graph (ac-count) showing the time before fiddling with the antenna cable and after. Quite a difference.