Re: Looking for a way to plot V_LSR vs intensity in GNU Radio

[Paul Boven]

Hi Ellie,

On 7/24/20 5:09 PM, Ellie White wrote:
> My question is this: do any of you know of a method -- or a GNU Radio 
> module -- compatible with GR version 3.8 that allows you to plot raw 
> data from a USRP and get a plot of intensity versus velocity with 
> respect to the local standard of rest (V_LSR) -- instead of intensity 
> versus frequency, as given by the time sink block?

We solved this somewhat differently: I wrote some Python code that, 
given Ra/Dec (which we can get from the telescope) and 
date/time/position, calculates the V_LSR for that viewing direction. The 
calculation itself is done in Astropy, hence the need for Python3, and 
the reason we upgraded to GR 3.8 as soon as we could. One of our other 
volunteers wrote a Telescope class that can be queried for pointing 
direction and the like.

The frequency offset is calculated every second, and sent back into the 
flowchart using ZMQ. The ZMQ sink sends the tuning commands to a 
frequency-translating-fir block to shift the received frequency 
automatically. The advantage of using this method is that the frequency 
bins for the FFT integration stay at the same V_LSR during long 
observations, or observations carried out at another date/time, so you 
don't need to resample/interpolate to compare.

As we show the integrated spectrum using the Vector Sink anyway, it's 
very easy to have it calculate the x-axis labels in terms of doppler 
velocity, instead of frequency.

We took some extra effort that the doppler shift calculation can simply 
be included in the flowchart, and starts/stops with the flowchart.

Happy to send you some code, although some of it is specific to the 
Dwingeloo telescope.

Regards, Paul Boven.