This recent video describes successful LoRa TTN data transmission from Picacho(*) Peak (elevation 3,370 feet), over 60 km to downtown Tucson, Arizona (elevation 2,200 feet). Looks like a fun climb, with a great view!
Meshtastic is an open source network alternative to LoRaWan or TTN, and the current range record is 254 km: a cross-border international Canada/U.S. achievement, with nodes in mountainous terrain! This long post describes the effort in detail.
Meshtastic is very interesting, but currently requires PlatformIO and Visual Studio, and runs only on a few select bits of hardware.
It would be fabulous if Meshtastic could be made to run on more popular and accessible platforms, like the Arduino IDE and Adafruit MCU/LoRa boards, but I suspect it would take quite a bit of effort to port the code.
(*) Aside: "picacho" is Spanish for "peak". Go figure.
That was a lucky day. RF range is not predictable even under the best conditions. There are so many variables that will have an effect on range. Just because they did it do not count on even 1/2 of it. What you want on the Arduino should be there for the looking. I have seen quite a bit of LoRa projects over the last few years. The range is not effected by the Arduino or other processor used it is the radio, receiver, antennas, placement and conditions that are the governing factor.
Good to hear that them across the pond have eventually discovered LoRa.
The distances reported are as expected really, all normal.
I have a nearby 40km hilltop to hilltop link across the Bristol Channel in South Wales.
Way back in time, 2014, when LoRa was little known, rather than just test for a 'record' I wanted to know how well LoRa would actually perform as in how much power was needed to achieve a link. So I used a decending power method of testing and with LoRa settings of Spreading Factor 8, bandwidth 41.7khz at 434Mhz using plain 1/4 wire antennas, the 40km link was covered with 3dBm of transmit power, the 40km was a LoRa land record at the time.
Based on those tests I suggested that the LoRa devices would be capable of circa 622km LOS at low data rates but a lot though this was a bit looney tunes, but it was 2015.
Also in 2014 there had been high altitude balloon flights in the UK using LoRa, ground to balloon distances of 400km were recorded.
Unfortunately, the Meshtastic code is an enormous jumble, with countless options and board/processor specific setup code. It is very difficult even to sort out what might be the minimum combination that might allow two MCUs+radios to talk to each other.
I did take a look at the GITHUB, and I got the impression that the code was not 'open source' in the sense that you could easily re-configure it, maybe that was the point.
Not sure (?) why a group of people would really want an encrypted message link between mobile phones that did not go through the mobile phone system itself..........
If the setup allowed for direct program applications as in sensor setups, in a similar way to TTN\LoRaWAN but without the gateways, that might be useful. Especially with 2.4Ghz LoRa which has many channels and few duty cycle restrictions.
As for the not on the Arduino IDE point, I have seen on the TTN\LoRaWAN forum quite a bit of comment that if you were a serious\professional software developer you would not be using the Arduino IDE.
I also noted that the records claimed appeared to just be point to point LoRa links, so no news here.
Useful if the mobile phone towers are down. Think survivalists!
After looking through the Meshtastic code, I noticed that they started with the Radiohead RHMesh code. I had not looked at that closely, and after doing so, decided that would be a much better starting point, than removing all the unnecessary frills from Meshtastic.
In my part of the world the cellular service only covers the metro areas and main roads. CB radio is borderline unusable due to the large number of toxic users.
A pre-shared key encrypted comms system gives you a private spam free messaging service. Useful for location/asset tracking of off road vehicles, farm and mining machinery together with short message communication. Aside from the initial hardware cost there are no ongoing costs.
You can talk to a local meshtatstic device via serial or bluetooth so effectively it's just another comms interface, like a cellular modem but with no telco contract.
Agreed. That code has been around a while so its going to be good for Arduino compatibility.
I took a look at the RadioHead code and it does appear to be not microcontroller\Arduino specific. Although it does appear to use an interrupt on the DIO0 pin to sense RX and TX done which a lot of Arduino libraries do not use by default. Could be some differances on some Arduinos maybe.
There is an RX timeout interrupt, but the SX127X does not provide that so it may be manually added.
LoRa sensor network data to cell phone or laptop via Bluetooth, and possibly, as mentioned by @mikb55, text message communications in regions where cell phone service is not availabe. Internet connectivity is of no interest to me.
Certainly requiring Internet connectivity as per TTN\LoRaWAN can be a problem or significant expense in out of the way places.
One benefit of the TTN\LoRaWAN approach for sensor networks is that the sensor can spend most of their time in deep sleep, so battery life can be in the years.
For a sensor Mesh the nodes would need to be in listen mode most all of the time so will be using circa 360mAhr minimum per day, not so good for battery use, implying that solar panels are needed.
Porting the mesh code across to the newer SX126X devices would halve the power needs for listen mode, but the API is quite different.
Useful figure, thanks! If you have looked at the Adafruit LoRa/MCU modules lately, they have built in Li battery chargers, complete with an enable input so that the entire module can be powered up and down with an autonomous power timer, if desired.
The receive current of an SX127X is in the region of 12-15mA, the SX126X uses DCDC conversion to power the low voltage RF stuff so the receive current is cut in half.
I have used one with a SEEED XIAO SAMD21 to build a TTN\LoRaWAN node. The combination worked well. I especially liked the sleep which would allow you to pause a running program down to a 5uA current. So much easier to work with than the 'deep sleep' of the ESP32s etc.
Have you noticed that RadioHead already has a driver for these devices? I'm currently learning how to use the RHMesh examples, and it appears that to switch between the SX127X (RH_RF95) and SX126X devices, one need only load the appropriate driver.