Ask Dr. SETI ®

I have a SpectraCyber receiver, and am using it with your Horn of Plenty antenna design, to learn the basics as I build up my understanding before I launch into a larger antenna with greater range (reach into the cosmos). I have been befuddled with the SpectraCyber's output . It took me quite a while to discover that the voltage output was not proportional to power but that the IF gain setting was apparently proportional to power. I think it is because of the existence of a "baseline" voltage that comes from the square-law detector upon which rides the signal from the antenna which includes of course LNA noise, spillover, cable noise etc.
 
I find that I need the absolute value of some voltage which cannot be gotten from the IF gain setting. The variation of the "baseline" with temperature just adds another undesirable variable. I would like a system that gives me a Signal = constant (POWER) rather that Signal = constant (POWER + B) where B=baseline. Is this possible?
 
James, SETI League member

The Doctor Responds:

As you probably already know, James, radio telescopes typically operate in one of three modes: radiometer, spectrometer, or interferometer. Each has a distinct purpose and design strategy. A radiometer measures (and is calibrated for) the total incident power collected within its design bandwidth. A spectrometer displays (and compares the amplitude of) multiple frequency bins within its bandwidth. And an interferometer uses interference patterns between multiple antennas to increase spatial resolution, independent of amplitude calibration.

I think part of your confusion has to do with the fact that the Spectra Cyber is a spectrometer, not a radiometer. Thus, it was never intended to provide absolute (calibrated) power measurements. It should, however, be possible to determine the difference (in dB) between various spectral components. If you have a good hot source (say, the Earth, at an estimated 290 K) and an equally good cold source (say, the Northern sky, estimated at around 10 K), then you can use their indications to crudely calibrate your system to interpolate the noise temperature of objects in between those figures. You can pretty much ignore antenna temperature, since a horn (unlike a dish) does not exhibit spill-over, and is neither over- nor under-illuminated. But, unless the detector is operated in the middle of its square law region, and the signal strength variations are small, this will result in merely relative, not absolute, measurements. After all, a spectrometer is not a (much more expensive) calibrated radiometer!