LAB-006 · equipment
The Receive Chain: What Sixmile Run Uses and Why
Dee Okafor ·
The receive chain at Sixmile Run runs from a prime-focus satellite dish on a concrete pier to a software-defined radio in the equipment room, with a cooled LNA and a hundred feet of low-loss coax in between. This is a walkthrough of each stage in signal-path order, with the specific choices we made and enough context to understand why.
Dish
Sixmile Run’s antenna is a repurposed prime-focus satellite dish, several-meter class, pulled off a defunct residential C-band installation. We rebuilt the mount from scratch, replaced the feedhorn assembly, and set it on a concrete pier. Not on Amazon, not on any product page we can link to: multi-meter prime-focus dishes are a local-pickup or classifieds purchase, and the freight cost on anything with real collecting area is prohibitive. Buy local, buy used, bring a truck.
The dish’s job is gain. At 1420 MHz, the hydrogen line signal from the galactic plane runs in the range of tens of Janskys for bright emission regions — weak enough that collecting area drives sensitivity more than anything in the electronics chain. Receiver upgrades will not compensate for a small aperture. At some point you are adding gain without improving the signal-to-noise ratio, and the place to fix that is earlier in the chain, not later.
To see what a given aperture buys you in gain and beamwidth, or to size a cantenna feed for a dish you’ve found, use the dish and horn calculator.
Feedline
The coax run from the feed to the equipment room is about 100 feet of LMR-400 equivalent cable. Feedline loss at 1420 MHz is not subtle. At 1420 MHz, LMR-400 equivalent loses about 4.9 dB over 100 feet (Times Microwave’s own LMR-400 datasheet gives 5.1 dB per 100 feet at 1500 MHz); RG-8X loses roughly 10 dB over the same run. That 5 dB difference is real, and it is why the LNA sits at the feed and the long run comes after it. We use MOOKEERF KMR400 with N-connectors, which terminates clean, measures well on a VNA, and has held up outdoors through three winters. It is not Times Microwave stock, and the listing is honest about that.
The rule: spend on feedline between the dish and the LNA. After the LNA, the signal is amplified and feedline loss matters much less. Before it, every fraction of a dB is permanent. The receive-chain noise calculator turns this into kelvin for your own cable lengths, and shows what the same parts do with the LNA at the wrong end.
LNA
The low-noise amplifier mounts at the feed, before the coax run. This is the most consequential purchase in the chain for anyone building from scratch.
For 1420 MHz work specifically, we use the Nooelec SAWbird+ H1. It’s a SAW-filtered LNA: the surface-acoustic-wave filter ahead of the amplification stage rejects out-of-band interference before it reaches the mixer. At L-band, strong nearby signals including navigation and cellular frequencies can saturate a wideband front end and raise the noise floor across the band. The SAWbird’s filter removes that before it’s a problem. Noise figure runs around 0.6 dB at the hydrogen line in our measurements. Bias-tee powered; if your SDR doesn’t support bias-tee, an inline injector is cheap and solves it.
If you’re building a more general receive chain before committing to the water hole specifically, the Nooelec LaNA is wideband (20 MHz to 4 GHz) and inexpensive enough to keep a spare on the shelf. No SAW filtering, so more exposure to strong in-band interference than the SAWbird. A reasonable step for general L-band exploration, but not a substitute for the SAWbird once you’ve confirmed your target.
SDR receiver
An RTL2832U-based dongle behind a good dish, good coax, and the SAWbird+ H1 will show you the galactic hydrogen line. The receiver matters less than the front-end stages, and most people learn this by upgrading the receiver and being disappointed.
The RTL-SDR Blog V3 is the right entry point. The R860 tuner is cleaner at L-band than the older R820T2 variants, the software support is the best available at this price tier, and at $37 it is cheap enough to kill without losing much. Its dynamic range is the ceiling that will eventually push you to a better receiver, but you won’t hit that ceiling until the rest of the chain is in reasonable shape first.
The Nooelec NESDR Smart XTR v2 HF Bundle is the practical next step: wider tuning range, bundled upconverter for HF coverage, same RTL2832U architecture but a better package. It is a better RTL-SDR, not a different class of instrument.
Software
SDR# on Windows, GQRX on Linux and Mac, and GNU Radio all work as starting points. For hydrogen-line spectral work specifically, what you want is a spectral power density plot integrated over time, not a real-time waterfall display. Most of the amateur hydrogen line pipelines we’re aware of use GNU Radio or a Python front end driving a custom spectral line averaging loop. The software layer is out of scope for this page, but one note: if you’re watching a waterfall and looking for 1420 MHz emission, you’re probably not looking at the right thing.
Entry-tier option
A working 1420 MHz receive chain for around $150: RTL-SDR Blog V3 ($37), Nooelec SAWbird+ H1 ($50), and a home-built pyramidal horn: foil-tape-lined foam board or thin sheet aluminum, with a quarter-wave probe in a small waveguide section at the throat. Materials and N-to-SMA adapters cover the rest. Don’t substitute an off-the-shelf UHF Yagi. The common ones are cut for 70 cm (about 430 MHz) and are not 1420 MHz antennas at all.
The horn will not give you the sensitivity of a prime-focus dish. Its effective aperture is small, and detecting faint galactic hydrogen at useful signal-to-noise ratios requires longer integration times than most people expect starting out. What it gives you: a real, working, directional signal chain you can operate without a permanent installation, a concrete picture of where your noise floor actually lives, and a basis for deciding whether to commit to finding a dish. That is the right first stage of this project. Nobody should start by sourcing a dish. Size the horn with the dish and horn calculator. A horn on a fixed mount will mostly be doing drift scans, so the Galactic-plane transit planner is worth opening before your first session.
Building one yourself
This page describes our chain. The step-by-step version for yours is a six-part series: the $150 receiver, the first observing session, choosing between a cantenna, horn and dish (which also qualifies what we say about aperture in the Dish section above), LNA placement and coax loss, reading your first spectrum, and measuring the Galactic rotation curve.
Further reading
The Radio Sky and How to Observe It by Jeff Lashley covers hydrogen-line reception and receiver chain design with enough practical detail to be genuinely useful. It assumes some prior RF comfort, and the hardware sections are current enough to apply. This is the closest published reference to what this page describes.
Is Anyone Out There? by Frank Drake and Dava Sobel covers how the water-hole band became SETI’s working consensus and the reasoning behind listening at 1420 MHz. The hardware sections are dated; the methodology sections are not.
The Code Book by Simon Singh is about cryptanalysis, not radio astronomy. It is on the list because the historical method it covers, frequency analysis, structural inference, testing a hypothesis against new material, transfers directly to translation work. Priya makes every new person working on the signal interpretation side read it before touching the translation notes.
The equipment and book links throughout this page are affiliate links. Walt keeps the hosting costs spreadsheet and the numbers are not comfortable. If you were going to buy any of this gear anyway, buying through those links offsets a portion of what this project costs to run.