I want to implement a digital filter and tuned amplifier for (39 KHz and 50 kHz) in DSP.
Tuned amps are a very narrow band, high gain amplifiers. They are practically filters.
As the circuit gets its input from the wheel sensor mounted at the tracks and is prone to a lot of noise,is it possible to accurately reproduce the signal using DSP?
Noise is not an unwanted signal.
We should look into whether it is really noise or not. Which parts contribute to that noise? What are the characteristics of this noise?
Any DSP could chop off the data, but can't create new data from nothingness. Therefore; the proper analog design is crucial for any sensor app.
As I am new to DSP,can anyone guide me on how to approach it?
I was thinking to use Intel cyclone 10 (FPGA with DSP blocks) to implement the design. I could start with Matlab and later convert the code to verilog.
1)Write sensor requirements down; dynamic range, offset, nonlinearity, noise, etc...
2)How can solve these problems?
For instance; It is difficult to make a Kalman filter in analog, so I should put this in digital.
Canceling offset in the analog world could solve my dynamic range issues, So I should put those in analog. (Or build a hybrid ?)
3)Prototype...