Building a Meshtastic network from Scratch:Installing Node 0

Introduction

In recent years, Meshtastic has been growing strongly in Argentina. What first seemed limited to small groups in major cities has gradually spread to new places, and today new nodes appear almost daily in different parts of the country.

At first, the first mesh networks emerged in the most populated urban centers, where it was easier for several users to be active and connect with each other. Over time, however, the idea expanded beyond those areas and began to spark interest in smaller cities and suburban environments as well.

This is exactly the case in Salsipuedes, the town where I live and work. It is a small city of just over 20, 000 people in central Argentina, located near Córdoba, one of the country’s major urban areas. Despite being close to a large city, its hilly landscape creates some of the usual challenges for connectivity, and there are still areas where communications are weak or unreliable.

Another important factor is that Salsipuedes still has a strong memory of critical events, such as the major flood of 2015 and, not long after, a serious wildfire in the surrounding area that had severe consequences for some neighborhoods. These events change the way people think about communications. In an emergency, having alternative, decentralized, and resilient systems is no longer just a technological curiosity. It can become an essential tool.

In that context, Meshtastic appears as a particularly interesting option. A network like this could provide an alternative communication path in areas with poor coverage or in situations where traditional systems fail, whether because of power outages, infrastructure problems, or disruptions to services such as the Internet or cellular networks.

By the end of last year, there were still no active public nodes in town, so I decided to start with the simplest step: running a few tests with portable nodes.

For that, I used two Wio Tracker L1 Prodevices and one SenseCAP Tracker T1000Efrom Seeed Studio to carry out range tests between different points in the city. Those first experiments showed quite clearly that Meshtastic had real potential as an alternative communication system in Salsipuedes.

After that experience, I began thinking about a more ambitious next step: installing fixed nodes that could serve as the foundation for a future local network. With that goal in mind, I started reaching out to different local institutions that could benefit from this kind of infrastructure, and that is how the SalsiMesh project began to take shape.

The proposal is based on the gradual deployment of nodes at strategic locations, with the goal of building a community mesh that can eventually provide coverage to a large part of the town. The project is still in its early stages and involves several local stakeholders. In future articles, I will share the progress, the tests, and the technical decisions that help shape this network.

For now, I want to tell the story of installing Node 0, the first concrete step in building this future local mesh.

Node 0

As a starting point, and with the goal of running some initial tests and gaining experience, I installed a fixed node at my home: SalsiMesh Node 0. The location is not exactly ideal. Although it sits noticeably higher than the surrounding neighborhood, it is still in a lower part of the city. Even so, I considered it a very good place to begin because having the node close to me makes it easier to try different configurations and run a series of tests to verify how it performs.

When choosing the hardware for this installation, I considered two main options: building a homemade node using a board based on ESP32 or nRF52840, along with batteries and a solar panel, or using a ready-made device. Since I wanted to avoid dealing with too many variables at this stage, I chose the second option and bought a SenseCAP Solar Node P1from Seeed Studio, a company whose products I know well and trust for their quality.

In the following sections, I will share my experience with this device and describe the installation, configuration, and testing process of this first node.

Preparation

Before installing the device in its final location, I had to complete a few preparation steps. The Seeed Studio wiki was very helpful throughout the process, since it explains everything step by step in a simple way.

Batteries

The first thing I did was install the batteries inside the device. The SenseCAP Solar supports four 18650 lithium batteries, which, at least in principle, should provide very good operating autonomy.

Ideally, it is best to use good-quality batteries from well-known brands, so the actual capacity matches the stated specifications. At the time of installation, I did not have batteries of that kind available, so I used other ones that had already performed well for me in previous projects. Time will tell whether they work just as well in this case.

I installed the batteries in the back of the device after removing several screws, making sure to insert them carefully with the correct polarity.

The batteries need to be the button-top type so they fit securely in the battery holders.

After installing the batteries and connecting the antenna, I tried turning the device on to make sure it was responding.

It is always important to power on radio equipment only after the antenna has been connected.

I had a brief moment of concern because nothing happened at first: the device would not turn on. But things made sense again when I checked the small instruction manual, which explains that the SenseCAP includes a protection circuit that enables the batteries only after the unit has been charged through the USB-C connector. Once I followed those instructions, the device started up without any problems.

Firmware

The device came with firmware already installed, but I decided to update it using the Meshtastic Web Flasher. After erasing the previous version, I installed the latest alpha release (2.7.20), an overly optimistic decision that caused some problems later.

Antenna

I used the antenna included with the device, which according to the specifications has a gain of 2 dBi. For extra protection, I wrapped the connector mounted on the bracket with a few layers of self-fusing tape to reduce the risk of water getting inside.

Configuration

With the device powered on and the firmware already installed, it was time to configure its different settings. This can be done either with the mobile app over Bluetooth or from a computer using a USB cable. I find the app more convenient and easier to use, and in my case I used the Android version.

The first setting I checked was Region, under the LoRa settings, which defines the radio operating frequency. In my case, that setting is ANZ (Australia, Brazil and New Zealand). On the same screen, I also adjusted the Preset, which sets several communication parameters as well.

After that, in the User settings, I changed the Long Name and Short Name that will identify this node.

Then I changed another important setting: the Role that this node will have within the mesh. In my case, this is a fixed node that will help build the network infrastructure, so I assigned it the ROUTER role. When we install other nodes, especially at higher locations that can provide wider coverage, I will review whether it makes sense to keep this role or switch to a different one.

Next, I also configured the Position settings for the node. This version of the device, the P1, does not include GPS, unlike the P1 Pro, so I entered a fixed position manually using the location provided by my phone:

Finally, I also configured MQTT access for one purpose only: to publish the node’s position on public maps, since I find that useful when planning the placement of future nodes.

In the MQTT section, I entered the server details for Liam Cottle’s map so the node could publish its position there. This server allows uplink packets but not downlink, so the node will not receive information from other nodes through MQTT.

I also configured packet upload and download for the LongFast channel, making sure that the node reports its exact position.

One thing I noticed while configuring the node with the app is that BLE (Bluetooth Low Energy) has a range of only a few meters. This is worth keeping in mind if you plan to place a node on a tall tower or in a remote location, because configuration can become more difficult. There is also the possibility of remote administration over LoRa, but that is something I plan to explore in the future.

Installation

I installed the node on a metal pole that was already mounted on the structure supporting the water storage tank. At one time, this pole held a TV antenna, and later it was used for a wireless internet antenna. Now that fiber optic service is available, it was no longer needed and found a new purpose as part of the Meshtastic mesh.

The node mount includes an adapter with a clamp designed for a pole about 10 cm in diameter. Since the pole I used is much thinner, I removed that adapter and used a smaller clamp directly instead.

The SenseCAP includes a 5 W solar panel to recharge the batteries. To achieve good performance, it is important to orient it correctly, so once the node was fixed to the pole, I took the necessary measurements to place it in the best possible position. The panel was oriented toward the north, as it should be in the southern hemisphere, and I set its tilt to about 31°, a value that at this latitude provides a reasonable balance for solar exposure throughout the year.

Testing

With the node already installed in its final position, I carried out a series of tests to verify that it was working properly and to check whether the configuration was appropriate.

Metrics

One node can request different types of metrics from another, providing very useful information. Since I have a couple of Wio Tracker L1 Pro devices, I paired one of them with my phone and, after selecting the SenseCAP from the node list, I used the Telemetry section to request Device Metrics and Signal Quality reports.

Device Metrics provide information such as battery charge percentage and battery voltage, as well as air utilization and channel utilization, which give an idea of how busy the radio channel is.

On the other hand, the Signal Quality report is useful for checking the RSSI values, which indicate signal strength, and the SNR, which reflects the signal-to-noise ratio seen by the mobile node. This also gives an indirect idea of how far the fixed node’s signal can reach.

Range Test

Another very useful test is the Range Test, which is specifically designed to determine how far the signal from the node under test can reach.

For this test, I also used two nodes: the fixed SenseCAP and the Wio Tracker as the mobile node. I enabled Range Test mode on both, but set a reporting interval on the fixed node, one minute, while leaving the mobile node without a reporting interval.

I left the “Save CSV to storage (ESP32)” option unchecked because this feature is only available on ESP32-based nodes, and both of mine use the nRF52840. With this setup, the mobile node stores the test results on the phone, and I can easily download them later.

Once the configuration was complete, the fixed node started transmitting a special message every minute with the content “seq x”, where x is a consecutive number that increases with each transmission.

Next, I went out around the area to check whether I was receiving the test messages or not.

One very important thing to keep in mind is that, unless you change the configuration, this test uses the main channel, LongFast. That means the test packets will be received by every node within range of the transmitter, which can be quite annoying for other users on the network. In my case, I was not too concerned because there are no other nearby nodes in my area, but it is definitely something worth considering before running the test.

After moving around the area near the node, I returned to the starting point and downloaded the log file with the test packets from the app:

The file is in CSV format (Comma-Separated Values). After transferring it to a PC, I opened it and confirmed that it contains, among other data, the receiver’s position and the SNR value measured at that point.

Although some conclusions can be drawn from the raw data alone, it is much better to analyze the results by viewing each recorded point on a map.

While searching the web, I found this Python tool from the YouTube channel The Comms Channel. It allows you to load the CSV file and display each recorded point, along with its SNR value, on different map layers.

Following the same idea, and after doing a bit of vibe coding with Gemini, I created my own version with some additional features, and it runs directly online with no installation required.

With this application, it is possible to view on the map all the points where a signal was received, with the corresponding SNR level shown in different colors. It is also possible to combine several CSV files, for example, collected on different days, and view them together or separately, filter by signal level, and see general statistical values. Also you can choose the interface language: Spanish or English.

This is the farthest point where I have confirmed that the range test signal reaches: nearly 3.8 km, with an SNR of -5.25 dB.

In another article, I will explain in more detail how to perform the range test using this tool.

Firmware Update

While running this last test, I noticed that I could no longer connect to the node over Bluetooth to make configuration changes. I tried resetting it, but the problem remained. This was probably caused by the firmware, since I was using a very recent alpha version, so I decided to downgrade to an older and more stable release, leaving it on version 2.7.15. After making that change, I did not run into the problem again.

Conclusions

The SenseCAP proved to be a very suitable device for installing a fixed Meshtastic node, since it includes almost everything needed for this kind of application: low power consumption, good autonomy thanks to its batteries, a generously sized solar panel, weather protection, and a structure that makes mounting easy in different situations. Its installation and configuration do not require special knowledge or skills, which makes it accessible to anyone who wants to deploy one and help expand the mesh.

My SAL0 node has now been running continuously for two weeks. During that time, it went through a few late-summer storms with wind and rain without any issues. Now, as autumn begins in this region, the amount of available sunlight is starting to decrease, which will give me a good opportunity to continue evaluating the system’s autonomy and overall behavior under less favorable conditions.

This is only the first step. SAL0 will be the starting point of a network that is just beginning to grow. In the coming weeks and months, we will add more SenseCAP devices as well as other nodes based on different boards, with the goal of expanding coverage and continuing to build SalsiMesh. As always, I will keep sharing the progress, the tests, and the lessons learned along the way.

Acknowledgments

I would like to publicly thank the Meshtastic Argentina community, whose members are always willing to answer questions, clear up doubts, and share their experience. Without their help, it would have been much harder for me to take my first steps and find my way in this fascinating Meshtastic world.

You can find the community on their website and Telegram channel.

If you have any questions or suggestions, feel free to leave them in the comment section below.

For more information and projects, you can check out my blog and social media.

See you next time! 🚀