Testing for low signal conditions

piotr337

I’d like to ask for advice on testing the feather in low signal conditions or total lack of signal. My code appears to work as expected in typical conditions but I need to start testing edge cases.

I connected a 30 dB attenuator and that didn’t appear to have an effect on the board.

Any advice is appreciated.

jaredwolff

You may need a larger attenuator @piotr337. I think the noise floor for the radio is quite low.

Another thing you can do without buying another attenuator is place the boards in something that will naturally attenuate your signal. Metal box, etc.

Another pro tip is to use the AT+CFUN=0 command to force disconnect your device, see how it handles it and then AT+CFUN=1 later on to make sure it connects again, etc. This definitely helped me work out some logic around disconnects/spotty connections.

AchimKraus

Metal box, etc.

Using a steel cooking pot with steel lid works also. There are some, which see that as capacitor antenna and you need to ground it with the pot, but in my experience it works also without.

piotr337

I’m having a difficult time getting the modem to fail.

I have wrapped the board with aluminum foil, put it a steel enclosure (Hoffman electrical enclosure). I grounded that enclosure to my power supply. I then wrapped the box with aluminum foil.

After all that the modem is still working as it did without all this metal surrounding it.

AchimKraus

Do you use NB-IoT or LTE-M?
Which signal strength does your device get outside and inside that “shield”?
I’m not radio-expert, but if the idea of a “capacitor antenna” is true, it may be
required to connect the antenna ground to the “steel case”.
That’s also required for that alu foil, it must contact the antenna ground
And if there is a gap in the shield, that’s also a gap for the radio.

Using LTE-M and a steel pot with closed steel lid throws my Thingy:91 “out of order”.
With NB-IoT the device still works “sometimes” even in that shield.