1. An Integrated 'Digital' Feed for 21cm Marcus Leech VE3MDL Science Radio Laboratories, Inc Image appears courtesy NRAO/AUI

2. Out shoutings Shouts out to: IRC gang Keenerd, DoYouKnow, mybit, LordKalma, atouk, RokLobsta, etc Mexican Crew: Deb 'n da chikkens Elvira and maybe Stan?

3. Presentation Overview Quicky introduction to Radio Astronomy Description of the device –Why? –Challenges and solutions A few preliminary results Video “tour” of the device

4. What is Radio Astronomy? Astronomy at wavelengths from a few mm to tens of meters Visible light has wavelengths in the region of 500nm, that is, 5.0x10 -7 meters From a physics standpoint, there's no difference between visible light, and microwave/radio-wave “light”. Living things have receptors for only a tiny part of the EM spectrum

5. Optical vs Radio Astronomy Ability to resolve fine detail highly dependent on wavelength A 10cm optical telescope can resolve details that would require a radio telescope over 42km in diameter at 21cm wavelength! Sensitivity, however, is proportional to collecting area of the reflector, regardless of wavelength  Both use parabolic reflectors  Both must have a surface that is within 1/10 th of a wavelength of a “perfect” parabola.

6. The Electromagnetic spectrum

7. History of Radio Astronomy Like much in science, it was discovered accidentally Karl Jansky, 1933, working on sources of static on international radio-telephone circuits at wavelengths of 10-20m. Discovered that static rose and fell with a period of 23 hours, 56 minutes.  Must be of celestial origin

8. History, continued Built directional antenna Pinpointed source at galactic centre, in Sagittarius

9. The Genesis of Radio Astronomy Science Jansky was re-assigned to other projects after his work on radio-telephone “hiss”. Several years went by with nobody understanding the significance of his discovery Grote Reber picked up on Janskys work in 1937, building a 30ft dish in his back yard.  Eventually mapped entire Milky Way emission at 160MHz (1.8m wavelength)‏  Published in Astrophysical Journal in 1944 Radio Astronomy now taken seriously

10. Grote Rebers Dish Now preserved as historical artefact at NRAO, Green Bank, West Virginia

11. Rebers observations 160 and 480MHz Skymap Made by hand from dozens of chart recordings

12. Radio Astronomy Today Radio Astronomy at the cutting-edge of astrophysical research  Roughly 70% of what we know today about the universe and its dynamics is due to radio astronomy observations, rather than optical observations Big projects all over the world  VLA, New Mexico  Arecibo, Puerto Rico  GBT, Green Bank, West Virginia  Westerbork, Jodrell Bank, ALMA, Hat Creek, SKA, etc Scientists named the basic flux unit after Karl Jansky  1 Jansky == 10 -26 watts/hz/meter 2

13. How does the cosmos broadcast? Multiple mechanisms for emissions  Blackbody radiation  Synchrotron radiation  Spectral lines from molecular and atomic gas clouds Universe is more of a chemical “soup” than you'd guess from optical observations alone. RA lets you “see” the invisible.  Pulsar emissions  Maser emissions Special case of molecular line emissions  Cosmic Microwave Background

14. Blackbody radiation All objects that are warmer than 0K emit EM radiation over a wide spectrum Warmer objects have higher peaks, at higher frequencies (shorter wavelengths)‏

15. Synchrotron radiation Charged particles (e.g. electrons) accelerating through a magnetic field Intensity higher at lower frequencies Above 1GHz, synchrotron radiation very weak

16. Spectral Line Emissions Many atomic and molecular species undergo emissions due to quantum phenomenon Emission is spectrally pure: emitted at discrete frequencies, rather than a range of frequencies Lots of really big gas clouds in interstellar space, and in star-forming regions within galaxies

17. The 21cm hydrogen line Emission at 21.11cm wavelength (1420.40575MHz). Van De Hulst proposed existence of neutral hydrogen in interstellar space in 1944. Successfully detected in 1951 by Ewen and Purcell at Harvard, using very modest instrument Confirmed weeks later by team in Netherlands headed by Jan Van Oort.

18. 21cm line continued Density of interstellar hydrogen very low  Less than 1 atom per cc of interstellar space! Emission caused by collisional energy transfer, causing electron spin change in neutral hydrogen A photon gets emitted at 21.11cm For a given atom, “perfect” collision only happens about once every 100,000 to 1,000,000 years! But along any given line of sight, there's a staggering amount of neutral hydrogen

19. Spectral lines and doppler effect Existence of spectral emissions allows science to map velocities of gas clouds within and outside the galaxy: thermal and rotation component. Doppler shift changes the observed wavelength/frequency of emission. Just like approaching/receding train whistle You can compute relative velocity by using the shifted wavelength and comparing to the “at rest” wavelength. EXTREMELY IMPORTANT RESULTS

20. An experimental 21cm feed Can an amateur-scale effort produce a “digitize at the feed” scheme for 21cm with acceptable performance? Can differential radiometry techniques, previously studied by myself and Ken Tapping at 70cm and HF, be used at 21cm? http://www.sbrac.org/files/DTP_RX.pdf http://www.sbrac.org/files/DTP_RX.pdf

21. System design: Sky side

22. System design: Reference side

23. System Design: Thermal Analog RF components and receivers are attached to a thermal slab: 20cm x 20cm x 3cm T6 6061 aluminum. LNAs firmly bolted to underside of slab, and covered. –Hopefully follow a common thermal “destiny” –Other parts also slab-attached: filters, line-amps, etc Compute module is passively cooled –NO FANS!

24. System Design: Shielding Slab forms top of shielded enclosure Walls of 1mm 3003 aluminum sheet Compute module and networking module attached to inside of walls Power filtered going in and coming out of shielded environment Ethernet connection via optical link Backplate of circular waveguide feed forms bottom of enclosure. Taped seams where necessary

25. System Design: Noise Source Noise source, controlled by Compute Module Provides roughly 6000K (13dB ENR) output CAL pulses every 30 minutes, for 30 seconds Simple 1N5235, 6.8V Zener. No amps required. –Lower noise levels required –30K increase when CAL fires –Roughly same order as sky noise at 21cm –Very stable and repeatable

26. System Design: Ref. termination Type-N 50-ohm terminator Connected to thermal slab Has extra thermal balast –Press fit into copper pipe –Copper pipe filled with pennies REF with CAL pulses: quantitative results

27. System Design: Compute Module Currently 1 st -gen Odroid C1 Quad-core S805 ARM-based SOC @ 1.5GHz 1GB memory USB, 1GiGe GPIO Passively cooled—bonded to skin and heatsink Powered with linear-regulator array –Cheap 5V switchers seem to fail at 0.5 rated current –Eliminate PS switching noise

28. Assembled System

29. System Design: Flow Graph!

30. Experimental setup

31. Preliminary Results: Spectral In my driveway, with feed extension

32. Preliminary Results: Spectral

33. Preliminary Results: Stability

34. Comparing Ewen/Purcell receiver They spent about $5000.00 (2015 projected $), mine is under $500.00 Aperture roughly 8 times larger than my “test” horn Integration time O(hours) vs O(seconds) –Tsys ~3500K vs Tsys ~100K Technology has come a long way, baby!

35. Next steps More driveway testing Test with optical ethernet –Still waiting on plumbing –Hopefully eliminate all spurs Upgrade to B210 (or maybe dual B200mini) receiver –Will required XU4 upgrade as well Build a “weather shell” for it, possibly with more shielding in the shell.

36. Questions?