Change your cover photo
Upload
vu3kqf@gmail.com
Change your cover photo
UG class of 2010, PG class of 2014, JE @ BSNL from 2014-2015, AP @ Dept of ECE, LBSCEK from 2015-2019, TGT WE @ KVS from 2019- till present.
Interest are MCU's, Audio, Radio etc.
This user account status is Approved

This user has not added any information to their profile yet.

Feather SDR -- A Software Defined Radio (SDR) using MAX78000
MAX78000 Feather Board

Feather SDR is a software defined redio receiver that can receive signals from 500KHz to 30MHz. It supports all demodulation schemes including AM, FM, CW, SSB with variable bandwidth selection etc so that the same receiver can be used with various digital modes like FT8, WSPR, DRM also.

So what is an SDR?
An SDR or software defined radio is very similar to an analog radio in operation. Both are able to receive the signals transmitted from a radio station. But the difference lies in the architecture inside. A conventional radio does all the processing in analog domain. It uses front end filters, rf amplifiers, mixers, local oscillator, IF filters, IF amplifiers, detectors for various modulation schemes, and finally audio amplifier for driving the speaker. The SDR architecture is entirely different. The signal from a preselector is given directly to a digital mixer along with a quadrature signal coming from a local oscillator which produces the quadrature IF signal. These signals are called I (in phase) and Q (quadrature ) signals. Now these I and Q goes to the stere microphone input of MAX78000 feather board. The rest of the processing is done inside the feather board and the audio comes out of the line out that drives a headphone directly.

Working principle.
The I &Q are first digitized into samples with a sampling rate of 48ksps. This ensures that a spectrum of 48khz wide is visible around the tuned frequency. Any signal that falls inside this bandwidth can be viewed in the display.
The I&Q are given to the MCU that works as the DSP here. Complex signal processing routines such as FFT, IFFT, Convolution, Hilbert transformation etc are implemented easily with the help of libraries. Advanced noise reduction techniques with learning capabilities can be implemented quite easily due to the processing power available in the MCU. Demodulating can be done very easily from these I&Q signals with simple math. Once the demodulated samples are obtained, it again goes into the codec chip for getting converted back to analog domain before going into the headphones.

Advantages of sdr over analog radios.
1. The flexibility and scalability stands first when we compare SDR with it's analog counterpart. You want to implement a new demodulation technology into your receiver, your analog receiver will make you sit with your soldering iron for hours along with pcb changes, new components etc. Whereas our sdr may require a software updating to do that. So basically no hardware changes are required.

2. Viewing the spectrum in realtime is the second biggest advantage when it comes to sdr's. You can always view the 48khz spectrum around your tuned signal in the spectrum scope. You can always see the activities in the band and can tune into a place where there is an activity going on.

3. Advanced learning algorithm for noise reduction are planned for this receiver. Normally used LMS weiner filtering can be replaced with much more adaptive techniques and routines since we have a powerful cortex m4 capable of doing machine learning also.

4. Another area where sdr stands apart are the effective implementation of AGC ( Automatic gain control ). This is particularly useful when we are trying to listen to a weak station surrounded by strong local stations. Several agc algorithms can be switched in software that best suits our situation. We can't have many such circuits in analog radios. ?

5. Typically in analog radios, we use mechanical filters which are difficult to tune and make due to the unavailability of fundamental crystals and crystal filters are very costly too. Ladder filters are a pain if we dont have the right equipments to measure and characterize each crystal we have.

Software filtering is very easy. You change the fikter coefficients in your code and bingo! You have a new filter with new characteristics. Typically each band and mode requires filters with different bandwidths which is also very easy to achieve with digital filters.