My motivation behind that is that we see more and more devices using wireless technology and it still remains pretty difficult to know what is shipped into the air. Of course you could use very expensive hardware, but the here, the goal is to make something affordable, something below 10$.
The first step when trying to understand a system is to know the frequency it uses. Nowadays, you can open the device and identify the RF chip. But sometimes, you are just unable to do it (nothing written on the chip, etc). But how can we know on which frequency it operates ? The idea here is to have a device that will try to determine a "known frequency". For example, a weather station will transmit the data at 315MHz. There are frequencies that are reserved for this type of use and for the ease of the design, will only assume the following frequencies: 315MHz, 390MHz, 434MHz.
To detect if there is a frequency in the air, we can use a logarithmic detector like the AD8313 from analog device.
The goal of this chip is to output a certain tension based on the input. Actually, they are essentially just cascaded amplifier. This is exactly what we are looking for: depending on the input it will capt in the air, the output will have more or less tension. However, this information is not enough. Indeed, you will have a high input if you are close to you weather station, but also if you are close to you Wifi station. In other words, the chip alone is not able to distinguish the difference in frequency.
To solve this issue, we will add different filters in front of the input, thus only the frequencies we are interested in will be detected. This brings us to a new hot topic: the filter design.