The hook
A whip antenna, an ESP32, and a LoRa radio the size of a matchbox — that's the entire receive chain standing between a rooftop in Junagadh and a satellite passing 500 km overhead. Before any of it goes on the roof, it needs a plan.
The problem
TinyGS makes the software side approachable, but a real station still comes down to five open questions: what hardware, what signal path, what it actually shows you once it's running, what to build first, and what "done" even looks like for a station that's never finished growing. This post walks through that plan, laid out before a single antenna goes up.
How TinyGS works
The signal path is short by design. A satellite in low Earth orbit transmits its downlink at 433 MHz; the antenna picks it up and hands it to an SX1278-based LoRa module (a Ra-02 in this build), which demodulates the spread-spectrum signal. From there, an ESP32 takes over station logic — tracking the satellite's orbit with SGP4 propagation, timing the uplink/downlink windows, and forwarding whatever gets decoded up to the TinyGS cloud, where it's stored, processed, and turned into something readable. During an active pass, packets typically arrive every few seconds, so the whole chain — antenna to cloud — has to keep up in near real time.
What TinyGS enables
Stripped down to fundamentals, a single station gives you four things: it receives signals, decodes telemetry, visualizes the resulting data, and — because the whole stack is open source and extensible — it lets you build on top of any of the other three. Decoding is the load-bearing piece of that list: everything downstream, from the dashboard to the pass history to the shared community map, depends on getting a clean decode first. Get that wrong and nothing else in the chain matters.
Hardware overview — the KS14NM build plan
The parts list for this build, and the reasoning behind each piece:
- 433 MHz whip antenna — ground-plane mounted, for RF capture at the satellite downlink frequency.
- SX1278 LoRa module (Ra-02) — handles spread-spectrum demodulation at 433 MHz; this is the component the rest of the decode chain depends on.
- ESP32 dev board — runs orbit tracking (SGP4), station logic, and the Wi-Fi uplink to TinyGS.
- OLED display — a local pass/status readout, so the station doesn't need a laptop tethered to it to know what it's doing.
- microSD card — logs raw packets locally, independent of whether the cloud uplink is working at that moment.
- Power supply — sized for continuous 24/7 operation, not just test passes.
- Weatherproof enclosure — the station lives on a rooftop, so this has to survive actual weather, not just a desk.
A few planning notes going into rev A, before any hardware gets mounted:
- Measure the noise floor first, before anything is permanently installed. A bad noise floor reading discovered after mounting is far more expensive to diagnose than one caught on the bench.
- Optimize power draw early. This station is meant to run continuously, which is a different design target than something that only needs to survive a single test pass.
- Mount on the rooftop, and when choosing exactly where, prioritize clear sky visibility over convenient cable runs. Cable loss is a solvable problem; blocked sky isn't.
What the live dashboard will show
Once the station is actually receiving, TinyGS's dashboard is the payoff:
signal strength over time, an orbit track plotted on a world map, current
pass information, the uplink/downlink frequencies in use, and a ground
station card reading KS14NM, Junagadh, India. The two numbers worth
watching most closely during any given pass are SNR — higher is
better — and, more basically, whether the pass completes cleanly at all.
That second one sounds trivial, but it's the actual bar for a first
successful contact. None of this is live yet; it's the target the rest of
this build plan is aimed at.
Key learning (so far)
The biggest design decision in this plan isn't the antenna choice or the radio module — it's sequencing. Noise floor measurement and power budgeting have to happen before the rooftop mount, not after, because both are dramatically harder to retest once the hardware is sealed inside a weatherproof enclosure three meters up. Deciding that order now, on paper, is a lot cheaper than re-learning it after an install.
The journey ahead
Roughly in order, here's where this build is headed:
- Build Logs — hardware assembly, enclosure, mounting
- Antenna Tests — gain, noise floor, feedline loss
- Telemetry Captures — first decoded packets, dashboard validation
- Satellite Passes — full pass reports once the station is live
- Failures & Fixes — because something will need fixing, and that's worth documenting as thoroughly as the wins
The plan from here is simply to document everything, including the parts that don't work the first time around.
Close
A home-built TinyGS station: tracking, listening, decoding. The sky isn't the limit — it's the start. First signal coming soon.