Signal strength jumps sporadically in rare cases

[jonashoechst]

In some cases we observe a plateau of increased signal strength of a single frequency / duration combination on a single receiver. The jump was observed in lab scenarios, as well as in the field with no users near the station.

[jonashoechst]

RTLSDR devices not only do have the gain control already configurable in pyradiotracking, but also another AGC which seems to be part of RTL2832. It has been reverse engineered about 10 years ago and is available in pyrtlsdr:

https://github.com/roger-/pyrtlsdr/blob/5b542375fa5c105f7c9fbb5016a557d27b5ad497/rtlsdr/rtlsdr.py#L374-L384

The details of the implementation are also discussed in this German article.

Initially we thought having it not disables could be a reason for the signal strength jumps, however it seems to be disabled by default. Explicitly disabling it in pyradiotracking did not change gain levels, however explicitly enabling it increased gain levels which leads to the assumption, that it is disabled by default.

[jonashoechst]

There do exist rtlsdr devices build with an E4000 aka. e4k tuner, which has 6 intermediate frequency (IF) gain stages, that can be used to specifically amplify the signal. A great explanation is given by @paultag on his blog.

However, the tuner which is currently regularly used with pyradiotracking is the Nooelec NESDR SMArt, which contains an R802T2 tuner.

[paultag]

Oh wow humbled you came across that and found it useful! Happy to share what I know in the course of playing with this all! It's very fiddly.

Additionally, this datasheet may help! https://www.nooelec.com/files/e4000datasheet.pdf

[jonashoechst]

Explicitly disabling the AGC was implemented in 5de9e6d.