Hello All. I only recently discovered the TI forums and they have been immensely helpful with some research that I'm doing. I'm looking to build a heterodyne interferometer that can not only measure sub-wavelength displacements, but perhaps Doppler as well. The basic operation is that we would have two sets of balanced detectors, the output of one pair would serve as reference beat while the other would contain the measured signal, which would have some phase and frequency shift as a function of Doppler and displacement.
I will send the detector outputs through a TIA and maybe one more amplifier stage before Band Pass Filtering. I'm considering using an IF frequency of between 10-20MHz. I would like to digitize the resulting measurement and reference signals and then digitally convert them to I and Q components before ultimately sending the data to Simulink for analysis.
I'm not sure if the best option is to use a DSP evaluation board. And if so, are they capable of sending the resulting data directly to a PC for Simulink analysis, or would I need to use a FPGA board as an intermediary? Could the FPGA itself be used to analyze the data? I also haven't completely specified the parameters of the system, but of course low noise and high SNR is important. So I'm thinking a DSP that is dual channel with about 16 bit resolution is in the neighborhood of what I would need.
The other important question is the sampling rate. If I choose to work at an IF of 20MHz and 3MHz BW to account for Doppler spread, by applying Nyquist I would expect to need a sampling rate of at least 45MSPS. But because the signal is IF, I also read that we can undersample using a rate based on the bandwidth of the signal rather than the highest frequency component.
Any advice that you could provide on ADC options, evaluation boards, etc would be appreciated. I'm more of an electromagnetics/analog guy, so the AD conversion is something of which I'm less comfortable. So I'm sorry if the questions are less than coherent.
Dave