Skip to content

Phase 2: unboxing in Garibaldi (Rio Grande do Sul)

Phase 2: unboxing in Garibaldi (Rio Grande do Sul)
November 7, 2025

Before taking new equipment to the forest, it pays to get to know it somewhere you can easily turn back. In November 2025, the TECER team spent a week in Garibaldi, in the hills of Rio Grande do Sul, unboxing the new generation of TVWS radios and the LoRa radios running Meshtastic. Vineyards took the place of the forest. The question was the same as in Phase 1: does the equipment work as promised?

Aerial view of a valley with vineyards and woods, a narrow road and a cluster of houses and sheds in the middle, under a blue sky with clouds
The valley of the test house, between vineyards and woods, seen by drone, Garibaldi (Rio Grande do Sul), November 2025.

Why test in Rio Grande do Sul?

One of the lessons of Phase 1 was to test first, in a controlled and easy-to-reach place. The test house, surrounded by vineyards and wooded slopes, had what fieldwork in the forest does not offer: stable power, internet and time to repeat the tests. The hilly terrain, with ridges and valleys, also made it possible to watch the signal crossing obstacles.

What was in the boxes?

The Phase 2 TVWS kit, bought from the manufacturer WiFROST, came with:

  • two base stations, the main radios the signal starts from;
  • two EPCs, small computers that act as the network core: they authenticate the CPEs and route data to the internet;
  • four CPEs, the smaller radios installed where the connection will be used;
  • sector antennas, brackets, cables, PoE power supplies and mounting accessories.

On the first day, 3 November, the team opened the boxes, checked the parts and powered up one base station and one CPE.

A CPE kit unpacked on bubble wrap: the white radio, metal mounting brackets, screws, a blue network cable, a black PoE power supply, a power cord, a plug adapter and a small antenna with a base
Parts of a freshly unpacked CPE: radio, brackets, PoE supply, cables and antenna, Garibaldi (Rio Grande do Sul), November 2025.

Over the following days, the radios were mounted on masts with the sector antennas, the cables were assembled and the test network was drawn on paper before being configured.

TVWS radio with a grey metal housing and heat sink, seen from the front, showing two antenna connectors, a network port and a power input
One of the new TVWS radios, with antenna connectors and a network port, Garibaldi (Rio Grande do Sul), November 2025.

How did the LoRa and Meshtastic tests go?

Alongside TVWS, the team flashed the Meshtastic firmware onto LILYGO and Heltec LoRa radios and built the first nodes of the network’s second layer. One of them was hooked up to a small solar panel.

The simplest test was also the most telling: driving along the dirt roads between the vineyards with a radio in hand to see how far the messages reached. On the screen, each message received shows which node it came from and how long ago.

A hand holding a LILYGO LoRa radio with an antenna and a small screen, inside a car, with a dirt road between grapevines blurred in the background
Meshtastic range test by car, among the vineyards, Garibaldi (Rio Grande do Sul), November 2025.

What did the measurements show?

With a handheld spectrum analyzer, the team checked that the TVWS radios were transmitting on the configured channels, in the UHF band used by television. It is the same care taken in Phase 1: before blaming the link, make sure the signal is on the air.

Screen of a handheld spectrum analyzer showing a signal peak and a reading of -48 dBm at 536.76 MHz
Reading of a TVWS channel on the handheld spectrum analyzer, Garibaldi (Rio Grande do Sul), November 2025.

The radio link came up. The CPEs connected to the base station and showed as registered in the manufacturer’s system. But data did not get through: a computer plugged into the CPE received a network address but could not browse. The behaviour repeated with all four CPEs and both base stations.

On 5 November, with the test window nearly over, the team opened a support channel with WiFROST. A technical meeting took place on 7 November. The manufacturer reviewed the settings and logs and pointed to the network feeding the lab, not the radios, as the likely cause. The diagnosis was not conclusive.

Where did the equipment go?

On 19 November, the kit was packed and shipped to Belém, in Pará, where the permanent installation will be. The working base in Pará is the Guamá Science and Technology Park (PCT Guamá), next to UFPA and LASSE, a partner lab since Phase 1. By late November, the team was already at the park, preparing the next stage.

Yellow façade of the Espaço Inovação building, with the park’s logo, under a blue sky, with an antenna tower on top
Espaço Inovação building at the Guamá Science and Technology Park, Belém (Pará), November 2025.

What we learned at this stage:

  • Testing far from the forest pays off. In one week the team got to know the equipment, built the network and found a problem that, in the field, would have cost a whole trip.
  • A radio link on the air is not the same as internet arriving. The signal can be perfect and data still not get through.
  • The LoRa layer is simple to build and test. A few radios and a small solar panel are enough to exchange messages without internet.

What comes next?

The data problem travelled with the equipment. The next stage was to investigate, on the bench, why the link came up but the internet did not arrive.


Further reading


Supported by: Internet Society Foundation (BOLT).

Last updated on