DIY Radio Telescope in a Shed: How I Caught a Galaxy's Signal

What does it take to hear the galaxy? As it turns out: a rusty satellite dish, a tin waveguide, an RTL-SDR dongle, and a lot of patience. This article documents my construction of a meridian-mount radio telescope tuned to the 21-centimeter hydrogen line at 1420.4 MHz — the universal frequency at which neutral hydrogen atoms throughout the cosmos emit radio waves.

Chapter 1: Construction

The Antenna

The foundation of the system is a salvaged 1.8-meter C-band satellite dish acquired for approximately 1,000 rubles. After cleaning off rust and repainting it, the dish got a second life as the parabolic reflector for a radio telescope. The choice of C-band was deliberate: these dishes are designed to collect weak signals, and their parabolic geometry focuses incoming radio waves efficiently onto a central feed.

The Feed Horn (Illuminator)

The most critical component is the custom-built circular waveguide — the "throat" of the telescope that actually receives the focused radio waves from the dish. I fabricated it from galvanized steel with these key parameters:

• Waveguide diameter (D): 15.6 cm — exceeds the critical diameter of ~12.4 cm required for TE₁₁ mode propagation at 1420 MHz
• Waveguide length (L): 28 cm — optimized through electromagnetic modeling
• Probe-to-short distance (l): ~8.3 cm — calculated as λg/4 for proper phase matching
• Choke ring: Creates a quarter-wavelength groove (~5.3 cm deep) that suppresses parasitic surface currents and improves impedance matching

The choke ring is what separates a functional feed horn from an amateur one. Without it, surface currents creep along the outside of the waveguide and interfere with the received signal. With it, the antenna's radiation pattern tightens up and sensitivity improves noticeably.

The Receiving Chain

Minimizing the signal path between the feed and the first amplifier is paramount — every centimeter of coaxial cable at these frequencies introduces loss. My solution was to mount both the low-noise amplifier (LNA) and the SDR receiver directly on the feed horn inside a weatherproof enclosure:

• LNA: Gallium arsenide amplifier (~500 rubles) — GaAs devices have superior noise figures at microwave frequencies compared to silicon alternatives
• SDR receiver: RTL-SDR Blog V3 clone (~2,000 rubles)

The total receiver chain, from the waveguide probe to the USB cable running into the house, is contained in one compact assembly bolted directly to the back of the dish's focal point.

The Mounting System

Rather than build a complex two-axis equatorial mount, I chose a meridian mount with single-axis declination control. The telescope is oriented east-west, and Earth's own rotation sweeps the beam through right ascension over the course of the day — essentially turning the planet itself into the drive mechanism.

For declination adjustment, a DC motor drives a car starter ring gear (salvaged from a Lada 2101) with a position sensor and limit switches for safety. This gives me full north-south coverage to track different galactic latitudes.

The Control Panel

Because a project like this deserves to look the part, I built an aesthetically designed control console with:

• Declination control panel with servo tracking capability
• Sidereal time clock (which gains ~4 minutes per day relative to solar time due to Earth's orbital motion)
• Power and mode selection switches
• Analog gauges for system status monitoring

Budget Summary

Total hardware cost: approximately 8,500 rubles (~$115 USD at time of writing):

• Antenna dish: 1,000 rubles
• LNA: 500 rubles
• RTL-SDR receiver: 2,000 rubles
• Materials, electronics, cables: ~5,000 rubles

Chapter 2: Software and Observation Methodology

The Challenge of Weak Signals

The energy arriving at my antenna from interstellar hydrogen clouds is vanishingly small — thousands of times weaker than the electromagnetic output of a mosquito in flight. You cannot simply point the dish at the sky and see a signal. It is buried deep beneath the receiver's thermal noise floor.

The solution is integration: record thousands of consecutive spectra and average them together. Random thermal noise fluctuates independently in each measurement and averages toward zero. A coherent, repeating signal from a fixed direction does not — it accumulates and gradually rises above the noise.

Software Tools

• SDR# (SDR Sharp): Primary receiver control and real-time spectral display
• IF Average Plugin: Custom accumulator that records and averages sequential spectra, building up the weak hydrogen signal over minutes of observation

Observation Procedure

1. Point the antenna toward the Milky Way
2. Set the center frequency to exactly 1420.4058 MHz with 2 MHz bandwidth
3. Activate the accumulation plugin
4. Wait and watch as the signal slowly emerges from the noise floor over several minutes

Chapter 3: Results — Hearing the Galaxy

First Detection

When I pointed the telescope toward the constellation Cygnus and let the accumulator run, something remarkable happened. Out of pure chaos — the flat white noise I had been staring at — a narrow spike materialized at approximately 1420 MHz. Radio echoes from trillions of neutral hydrogen atoms in our galactic arm, each one emitting its tiny quantum of radio energy, adding up to a detectable signal in my backyard shed.

A Multi-Component Signal

Rather than a single clean peak, the spectrum showed three distinct signals at slightly different frequencies. This is the Doppler effect in action: hydrogen clouds moving toward or away from us at different velocities emit at slightly shifted frequencies. The three components correspond to:

• The strongest signal: nearby interstellar hydrogen in the local spiral arm
• A weaker signal: a distant spiral arm cloud approaching us
• Another weaker signal: a cloud receding from us at a different galactic distance

Galactic Tomography

This multi-component structure is the key to mapping the galaxy. By scanning systematically along the galactic plane and measuring the Doppler-shifted frequency profiles, one can:

1. Determine cloud velocities from frequency displacement
2. Estimate distances using galactic rotation models
3. Map the distribution of hydrogen across the spiral arm structure
4. Eventually construct a galactic rotation curve — which, at large radii, reveals the gravitational influence of dark matter

Future Directions

The telescope's current angular resolution of approximately 5° limits the detail of any galactic map I can produce. Two paths exist for improvement:

• Larger dish: A bigger aperture directly improves angular resolution, though it means a larger, more expensive, harder-to-mount antenna
• Radio interferometer: Two or more smaller antennas separated by a long baseline can achieve the resolution of a dish as large as their separation — the technique used by professional radio observatories worldwide

Fundamental science isn't somewhere else, in a large institution with a large budget. It is here, in a shed, in a tin bucket and a spike on a screen. It begins with curiosity and ends in a personal discovery of the Universe.