[Discuss-gnuradio] 21cm spectra after adding the front end and another line amplifier

[Marcus Leech]

Last week, I observed that with the USRP/DBS_RX combination terminated 
at the input, there was still
 a couple of spectral spurs when I set the system up to observe at the 
neutral hydrogen line at
 1420.405Mhz.  This worried me a little, but I surmised that once I had 
the LNA and another
 line-amp on, their added noise would swamp any little spurs.

Here's a screenshot of my RA application with the feedhorn setup on a 
1.25M dish, with a 15dBg/0.7dBNF LNA,
 a 15dBg/3.0dNF line amplifier, and roughly 6dB of cable loss to the 
DBS_RX front-end.  The spectrum
 is nice and quiet, which is just what you want, with perhaps the 
occasional HI region passing through
 the beam to de-flatten the spectrum :-)

I measured the response of the setup to a blackbody source at 315K this 
morning, which produced a
 massive (at least in RA terms) 2.25dB increase in the total power 
reading.  That's a ratio of
 roughly 56 (17.4dB) between the average magnitude of the random 
fluctuations seen in the total
 power display, and that produced by a blackbody source at 310K.   It 
could stand a further increase
 in front-end gain, I suspect.

But the main thing is I was very pleased to see the spurs go away, and 
also to see how relatively
 stable the system is.   I'll be making some attempts to capture 
transits of Cass. A--a standard
 calibration source that produces a flux density of about 2000Jy at the 
21cm wavelength.
 If that is unambiguously successful (and in theory, it should be, but 
in amateur observatories,
 there's nearly-always a big gap between theory and practice), I'll let 
the list know.

A couple of good messages to tak e away from this are that GnuRadio 
forms a good basis for
 amateur radio astronomy observing, and that USRP/DBS_RX combination 
make this possible
 with a bare minimum of hardware futzing required, and at a fraction of 
the cost of
 the only comparable amateur RA system on the market today.