Thank Alex for your answer, For the second question, I will not design a pulse doppler radar with a linear FM pulse. But I wonder that I could generate an LFM signal (to measure another RF device at 1GHz) using Ettus B200?
Due to the limit of the hardware capabilities, I think I can only generate an LFM with a maximum ~ 56 MHz sweep bandwidth (if possible). My approach is the modification of the source code of the FMCW block using OutOfTreeModules with gr_modtool, In which I will add the deadtime at the end of each FMCW pulse but I think I will not trade with my time to do this problem because I have already another solution :D Again, I appreciate your help and time in advance Vào Th 5, 8 thg 7, 2021 vào lúc 16:16 Alex Batts <battsale...@gmail.com> đã viết: > Hello, > > It has been a while since I used the FMCW block in the Radar toolbox, but > from what I can remember, the three samples sections are defining the > percentage of the signal you want distributed in each of those sections. > For example, if you want 50% of your signal to be CW, and 50% to be > up-chirp, put an equal number of samples in the Samples CW and Samples > up-chirp sections. The number of samples combined with your sample rate and > frequency sweep will determine the slope of the fm chirp. > > There is also a VCO block I'm pretty sure that you can use for more > customizable fmcw radar. > > For your second question, why do you want to add a dead time? My guess > would be because you are trying to do pulse doppler of some sort, but if > not please explain because I would definitely be interested in knowing what > you are attempting to do! If you are trying to do a pulse doppler radar > with a linear fm pulse compression, unfortunately, there isn't much > functionality within GNU Radio for that kind of thing, and SDRs aren't > great for pulse doppler right now. In a few years' time they will, but most > affordable SDRs just don't have the hardware capabilities necessary. FMCW > is the way to go for SDRs because of low peak power needed and because > timing is not as necessary for range disambiguation, which is nice because > of the low sample rates of most SDRs. > > I am not sure you can implement a dead-time within the FMCW signal > generator block, and am not sure that gr-radar has that capability. > However, if you still want to implement a dead-time, you could write a > custom block (GNU Radio has extensive tutorials on how to do this, and only > requires basic knowledge of yaml and c++). I would do something like, > generate a sine wave, and while the sine wave is above a certain threshold, > send the fmcw signal through, else, send a constant 0 through. You would > have to play around with the threshold to get the proper duty cycle and > play around with the frequency of the sine wave to make sure you are > staying in sync with the signal generator, but that is one idea on how I > would do it off the top of my head. > > Can you give me more information on the project and/or what your goals > are? It would help me give more specific advice as opposed to shooting > blindly. > > Thank you, > > Alex > > On Thu, Jul 8, 2021 at 5:44 AM Pham Van Dung <dungdk...@gmail.com> wrote: > >> Hello all, >> I want to make an FMCW signal generation with GNU Radar toolbox and Ettus >> B200 >> I saw that you have several examples in >> https://github.com/kit-cel/gr-radar/tree/master/examples/simulation, but >> I do not understand the following parameters of the FMCW block >> - Samples CW >> - Samples up-chirp >> - Samples down-chirp >> - Frequency CW >> Second question, if I want to modify the block (I want to add dead-time >> at each pulse of FMCW), How can I do it? >> I appreciate your help and time >> >> -- ----------------------------------------------------------------------- Van-Dung,PHAM
