LoRa or wifi ..... max received signal strength -20bBm?

ADR function.

Hi thanks for your replies.

Yes but WHY -20dBm ?
What scientific principle limits receivers to only report -20dBm no matter how strong the field strength?
I just tested my mobile phone running 'wifi analyser' app. near my wifi hub and from about 40cm to right against the hub the signal never wavered from -25dBm. A normal CW signal would have varied logarithmically as expected according to the laws of physics (ignoring multipath!)
What anomaly of physics are these devices employing to "defy the laws of physics Jim"?
Joking aside, I know -20dBm is recognised as a strong signal, but why -20dBm and not 0dBm or +20dBm.
I have mostly measured RF power referenced to zero dBm so what's going on?
Dave

There's no reason for it to be -20 dBm, any more than 0 dBm. Zero isn't a reference value, it's just a power level that happens to be written as a zero. You noticed your WiFi signal was -25, so that's a different value too.

It's too simplistic to try to match it to a log rule not just because of reflections, but also because the antenna detunes itself when it's held extremely close. It depends on the antenna design. Plus WiFi is a protocol, it's not just an unmodulated CW signal, and signal level might adapt (I'm no WiFi expert).

One reason why the reporting could be a round number (instead of say -21.5 dBm) is that the chip or driver or software might not be reporting it with any better granularity. 

Here's a screenshot from my WiFi access point, with a mobile phone held close to it. The value is -16, again no reason for it to be specifically -20 or -25 or 0. 

If the aim is to check the antenna quality, it won't be any good to do it so close up. You'd either have to accept the manufacturer's spec (assuming it is attached correctly) or try to do partial measurements to ensure it is matched as best as you can make it (which again doesn't give you any specific signal strength, since all bets are off when you're measuring so close) or you could try to do a DIY field test of some sort (because full antenna testing costs five-figure sums, at least).

Just guessing here, but -20dB in radio is 100uV.  Maybe there's a clamping circuit in the chip to keep the front end from being overloaded.  With it being consistent, it's got to be a feature - not a coincidence.

Just guessing here, but -20dB in radio is 100uV.  Maybe there's a clamping circuit in the chip to keep the front end from being overloaded.  With it being consistent, it's got to be a feature - not a coincidence.

Also some of the radio chips are so complicated, that it's easy to wrongly compute the measurement, or display a measurement that has not been adjusted to compensate for all the settings.

I don't know what LoRa chip is being used, but the (extremely popular) SX1276 has (seemingly unresolved) comments on the Internet concerning how to translate register values into a dB value. Some people reporting discrepancies in the multiple tens of dB. For that chip at least, the formula depends on if it's LoRa or FSK (the chip can perform both options), and on the SNR value reported in a different register, and whether per-packet measurements are being used (which for some reason are inaccurate) or a non-packet measurement (yet again, a different register for that), plus it's dependent on an amplifier gain setting. Plus there's an error in linearity noticeably at large signal levels for that chip apparently. But I don't know what chip is being used, nor what software, so someone might need to also do that investigation, and perhaps change register values to see the effect, if it was really important to get to the bottom of it. And doing all that, the value is still meaningless if the two antennas are right next to each other.