Re: Doppler

[Daniel Estévez]

On 01/01/2024 07:51, Jiya Johnson wrote:
> Yes I want to use 10GHz/s

Hi all,

Besides the physical considerations that Marcus and Marcus have 
mentioned, I *think* it should be possible to adapt the flowgraph you 
attached in your first message to generate correctly a chirp waveform at 
10GHz/s as you could have in some radars. (Here the "*think*" comes 
because most of these blocks use float parameters rather than double, 
which for some specific applications can be insufficient. Here I think 
this won't be a problem, but I'm not sure until I see it working).

The thing you need to do is to calculate some numbers and use a sensible 
sample rate for what you want (there is no reasonable way you're going 
to get a 10 GHz/s sweep in a 1 ksps sample rate, which is what you have 
in your flowgraph). For instance, say that your chirp waveform sweeps a 
bandwidth of 100 MHz. At 10 GHz/s, it takes 10 ms to sweep 100 MHz. This 
gives the period of your sawtooth. Let's say you use a sample rate of 
200 Msps, which is good to represent a 100 MHz bandwidth signal with IQ 
sampling.

Then you need to set the frequency of the sawtooth to 100 Hz ( = 1 / 10 
ms). If you want the sawtooth output to have units of Hz, its amplitude 
should be 100e6 (to achieve a 100 MHz sweep).

The formulas in the flowgraph you have all work correctly to this 
purpose, but to use them, you nee to set the drift_rate to 10e9 (10 
GHz/s), the drift_duration to 10e-3 (10 ms, see above), and the sample 
rate needs to be at least somewhat higher than 100e6 (say 200e6).

Best,
Daniel.

OpenPGP_signature.asc
Description: OpenPGP digital signature