ETA: Antenna Direct Sampling Experiment @ PARI: Nov 21, 2005

ETA: Antenna Direct Sampling @ PARI: Nov 21, 2005

Overview: Shown here are the results of testing the feasibility of direct sampling of the ETA dipole signals, as is planned for the completed system. It is found to be practical with 12 bits at 104 MSPS, even when the total signal is RFI dominated.

Maintainer: Steve Ellingson (Virginia Tech)
This page is http://www.ece.vt.edu/swe/eta/AB_051121/
Updates:
08 Dec 2005 UT 2130. Initial posting.

Measurement Procedure: One of the two dipoles at Stand 12 (the east outrigger) was used. This is a "production" dipole with a "production" active balun, connected to the electronics building by about 150 m of LMR400 coaxial cable (buried). Inside the electronics building is a rudimentary receiver consisting of some Mini-Circuits ZJL-3G amplifiers and SLP-50 low-pass filters. The output is input to an MCMS receiver which has modified to route its input directly to the input of its 104 MSPS, 12 bit A/D. The output of this A/D is captured in 16K blocks and recorded.

Full Span: Here we show the digitized result, obtained by averaging the power spectra of the 16K sample blocks, obtained using 16K FFTs with Bartlett window. The effective integration time is 418 ms and the spectral resolution is about 6 kHz. The vertical axis is shown in units of dB relative to the estimated temperature of the active balun (250K). Also shown (in red) is a prediction of the power spectrum of the Galactic noise background, obtained by combining Cane's empirical result with an electromagnetic model of the dipole response. The good news is that the agreement is very good, and the spectrum in the ETA band of interest looks possibly useable. The bad news is that we have picked up a 1 dB ripple. The period is consistent with a reflection over the length of the 150 m cable, so this is almost certainly not related to the direct sampling.

ETA Search Range (29-47 MHz): Here we are simply zooming in to take a closer look at the ETA band of interest. Note that most of the spectrum appears to be available.

Zooming In Further: Here we are simply zooming in to take a closer look at one small portion of the spectrum. However, note the units are now linear (not dB). This view contains the 37.50-38.25 MHz U.S. Radio Astronomy band. While there is no obvious RFI, neither is it clear that this would remain true if we continued to integrate. Also, while this result is typical, we do intermittently see interference in this band, and sometimes it is extremely strong.

	Full Span: Here we show the digitized result, obtained by averaging the power spectra of the 16K sample blocks, obtained using 16K FFTs with Bartlett window. The effective integration time is 418 ms and the spectral resolution is about 6 kHz. The vertical axis is shown in units of dB relative to the estimated temperature of the active balun (250K). Also shown (in red) is a prediction of the power spectrum of the Galactic noise background, obtained by combining Cane's empirical result with an electromagnetic model of the dipole response. The good news is that the agreement is very good, and the spectrum in the ETA band of interest looks possibly useable. The bad news is that we have picked up a 1 dB ripple. The period is consistent with a reflection over the length of the 150 m cable, so this is almost certainly not related to the direct sampling.
	ETA Search Range (29-47 MHz): Here we are simply zooming in to take a closer look at the ETA band of interest. Note that most of the spectrum appears to be available.
	Zooming In Further: Here we are simply zooming in to take a closer look at one small portion of the spectrum. However, note the units are now linear (not dB). This view contains the 37.50-38.25 MHz U.S. Radio Astronomy band. While there is no obvious RFI, neither is it clear that this would remain true if we continued to integrate. Also, while this result is typical, we do intermittently see interference in this band, and sometimes it is extremely strong.