SX1262 LoRa Connection Tester #1593
Replies: 5 comments 2 replies
|
LoRa is a lot of fun - yes! I don't see any mention of your antenna situation; of course I'm stating the obvious that few things are as important to RF comms as the antenna. It might help narrow & improve the range of responses you get from people who are super-knowledgeable w/ LoRa if you spelled out the type & positioning of your test antennas. |
|
Each module has a 5dBi stick antenna (I don't know what they're actually called), in λ/2, positioned vertically. You cand see them in the photo I included in the project description. Or here: https://www.aliexpress.com/item/1005006833587735.html . The whole setup of both modules it is placed on top of a stick that is held about 2 m off the ground. If you access the Web Interface I made for the tester, you can use the LoRa calculator I created, which also takes into account the antenna gains and other loses... There is more info on my project page (the first link). |
|
@MarusGradinaru As compared to other product data, 5 dBi seems to be rather optimistic for this kind of antenna. |
|
@MarusGradinaru it's great to see you running the experiments!
I would be rather sceptical about those range numbers. 500 kHz bandwidth combined with very low SF values is the absolute fastest that the LoRa radio can go, and probably only works for direct, line-of-sight connection in low-noise environment. So for longer ranges, you will definitely need to decrease the bandwidth and increase the spreading factor.
I don't know how the calculator is implemented, but that would be my guess as well, since modelling propagation through an urban/suburban environment is a lot more complex than just calculating the free space path loss (which is a rather straightforward formula).
Those are two different things: the noise floor is property of the local environment - the amount of RF power you receive when there is no local transmitter present. Sensitivity on the other hand is the absolute minimum power the radio needs to have on its input to receive the packet.
That is mostly correct, though if you want to get the noise floor you should run the
All of those things will definitely add up and impact the receiver sensitivity, though it is hard to guess by how much. The 5V pin on ESP32 is typically hooked up directly to the USB 5V pin, which is a pretty noisy power source. |
|
I have similar experience with poor range while trying out sx1262 + radiolib. |
I am currently working with the SX1262 Evaluation Kit and am facing an issue with the communication range. Our hardware is designed based on the Semtech reference circuit. During testing in an open environment, we observed that the communication link is lost when the two modules are separated by only about 10 steps, which is much lower than the expected range. At a distance of approximately 10 steps, the measured RSSI is around -121 dBm. Has anyone encountered a similar issue or have any suggestions on how to troubleshoot or resolve this problem? Any help would be greatly appreciated. Thank you. Referance Design : Thank you. |
|
Does this only happen after 10 steps ? At longer distances, does the RSSI improve ? |
Uh oh!
There was an error while loading. Please reload this page.
Hi !
I think many of you here are already familiar with this technology, but if you are a beginer in LoRa stuff, like I was 3 months ago, you can try this LoRa "Hello World!" project I made: https://github.com/MarusGradinaru/LoRa-Test 😃 For me, LoRa was just a dream 3 months ago... It was something so complex that I thought I would never understand how to use it. But my fear started to fade away when I began to study... And now I can't believe I managed to make my first LoRa connection. I am passionate about electronics and programming, and this technology will open up new horizons for me that I could not even imagine until recently. Thank you, @jgromes and all of you who contributed to this RadioLib work of art !
Fun fact: I was learning C++ and JavaScript for the first time too, while I was developing this project. 😁
Now I want to ask, those of you who know more about LoRa, a few things about these tests I made...
What my tests confirm is that the link indeed becomes impossible once the Link Budget is exhausted and the sensitivity threshold is reached. What doesn’t match, however, are Semtech’s range estimates (the last column). In my case, the actual range is much shorter: when they say 700 m, I already hit the Link Budget limit at around 250 m.
Could it really be that I have that much path loss in my rural environment ? As you can see in the picture, the main obstacles are mostly vegetation (trees), a small wooden cabin (with a metal roof, admittedly), a greenhouse, and that’s about it. The thing is, after about 250 m, the terrain starts sloping down into a small valley. Could all this really reduce the range so much ?
If that’s the case, does it mean that Semtech’s calculator assumes free-space line-of-sight conditions ? But if so, there’s already another formula for that — FSPL — which gives me hundreds of kilometers. So to me, it doesn’t really add up...
And one more thing...
Is it correct to assume that the RSSI reported by the SX1262 while idle (
.getRSSI(false)) represents the actual noise floor, against which the chip calculates the negative SNR obtained after despreading ?Because it seems that this noise floor measured in reality on the module is higher than the theoretically calculated one and presented in the datasheet, with +6dB added as internal noise of the chip...
In my setup I get 5-10 dB more, above what is presented in the chip datasheet... Could this be because of the noise produced by the ESP32 ? Especially since I also have WiFi active... The SX1262 module si powered from the +5V pin on the ESP32-DevKitC board.
All reactions