logo elektroda
logo elektroda
X
logo elektroda
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
📢 Listen (AI):
  • Close-up of several SX1276 radio modules packaged in plastic bags. .

    Some time ago I ordered some radio modules with the SX1276 chip for experimentation. After some time, I finally decided to get around to testing them.

    Battery from a disposable e-cigarette with an electronic circuit on a wooden background.

    In order to make a sensible estimate of the range and capabilities of these modules, I built several devices: two serving as receivers/transmitters and three transmitters. I programmed all the devices myself from scratch, avoiding off-the-shelf solutions such as Meshtastic.
    The receiving part

    Yagi antenna and open enclosure with electronic circuit on a wooden floor. .

    The receiving part consists of an SX1276 chip connected to an ESP32 on a WT32-ETH01 module and a DC/DC converter powering the whole thing from a 24V passive PoE adapter. The software receives and decodes the signal and then transmits it to the server, creating something like a 'LoRa gateway'. I have built two such gateways: one is connected to a Diamond X30 vertical antenna and the other to an ATK-5 dipole antenna in horizontal polarisation. Both antennas are mounted on a 55-metre stadium light pole, although the location is not ideal as it is at the lowest part of town, right next to the lake. I used about 2 metres of H155 cable at the antennas. In addition to receiving data from the ether, the software on the 'gateway' also allows transmitting and changing the radio configuration. In addition, it has a simple web interface for monitoring the status of the device.

    Screenshot of the ESP_LoRa GateWay interface displaying module parameters and information. .

    View of a lake with a radio antenna in the foreground.


    View of radio antennas mounted on a pole above a lake. .

    Transmitters

    Sealed box with antenna, used as a radio transmitter.

    Open electrical box with electronic modules and wires.

    The first transceiver created to estimate the potential was a locator built with ESP8266, SX1276 and NEO6P GPS module. The software in this device first waits for a GPS signal and then formats it into a string such as: "51.16960*69.95865*0" (position and speed). The entire string is encoded with the XOR function, a checksum calculated from the ASCII codes of the entire string is added, and then the whole thing is sent over the radio. Depending on the speed read from the GPS, the device goes to sleep for a certain amount of time. If the speed exceeds 30 km/h, the device does not go to sleep, but sends a position every 5 seconds. The time it takes to send such a string depends on the radio's configuration. I placed the whole thing in an electrical box, which, although it does not look aesthetically pleasing, works properly. The device powered by three 18650 cells works for 4 days. The first tests showed the great potential of this solution, as illustrated by the attached range map.

    Coverage map of radio devices around Ostróda. Map showing radio signal paths around Lake Drwęckie.

    After determining the range, I built another transmitter with ESP8266, SX1276 and DS18B20 sensor, placing it in an airtight box along with two 3.6V 17Ah lithium batteries. The device only reads the temperature from the DS18B20 sensor, formatting it in string: "L01|25.5|3.567" (ID + temperature reading + battery level), then encrypts and adds a checksum similar to the locator. I installed the device at a distance of 6.5 km from the receiver, in the middle of a forest near Ostróda, attached to a tree. The antenna at the transmitters is an ordinary 'stick' bought for £20. Experiments with a wire dipole gave similar results.

    Saft LSH 20 lithium batteries 3.6V in plastic packaging .

    I am very pleased with the results - using a simple method, I have achieved 'telemetry' with a range of practically the entire city, and at distances of up to 2 km, the signal even reaches the basements of buildings. This opens up the possibility of building an energy-efficient burglary radio system without subscriptions or other charges.

    In addition, I am measuring the radio background on the frequency 434.250 MHz, on which I am currently experimenting. I am attaching the temperature transmitter code, maybe someone will have interesting suggestions.

    Noise level graph for X30 and ATK-5 antennas. .

    Well and a temperature measurement from the forest :) .
    Graph depicting external temperature variations over several days.

    Cool? Ranking DIY
    About Author
    dktr
    Level 25  
    Offline 
    dktr wrote 878 posts with rating 651, helped 42 times. Live in city Ostróda. Been with us since 2003 year.
  • ADVERTISEMENT
  • #2 21152356
    gulson
    System Administrator
    Thanks for sharing your experiments! My modules are still lying ;) .
    I didn't expect such ranges.
    Text me a Parcel Post and I'll send a little gift! :)
  • #3 21152451
    TvWidget
    Level 38  
    dktr wrote:
    This opens up the possibility of building an energy-efficient burglary radio system without subscriptions or other charges.

    Remember the legally permissible bandwidth occupancy rate. Long range is achieved at the expense of low bandwidth. That is, transmitting a frame requires the transmitter to be on for quite a long time.
    You can occasionally send an alarm message but temperature monitoring is already problematic.
  • #4 21152460
    TechEkspert
    Editor
    Very interesting experiments, what is the purpose of xorating the coordinates?
  • ADVERTISEMENT
  • #6 21152525
    dktr
    Level 25  
    >>21152451 Yes, I know about this. With current radio settings it takes about 0.6 sec to send 30 bytes.
    >>21152460 Basically none, I experimented and forgot about it and it stayed that way ;)
    >>21152469 Lithium batteries (not rechargeable) 3.6 V 17 Ah.
  • #7 21152548
    acctr
    Level 38  
    dktr wrote:
    >21152469 Lithium batteries (not rechargeable batteries) 3.6 V 17 Ah.

    Is it worth using these instead of rechargeable batteries e.g. 18650? As I look the price of these batteries is around 70£ per cell.
    Helpful post? Buy me a coffee.
  • #8 21152748
    dktr
    Level 25  
    >>21152548 Yes, 17 Ah would be about 6 18650s, and here I have the same in R20 battery volume - plus no problem in sub-zero temperatures. The sensor draws 39 μA when asleep, 130 mA when transmitting and lasts 0.4s. According to estimates it should last about 7 years.
  • #9 21153675
    TechEkspert
    Editor
    "Energy density" is sizable, similar ones are found in disposable ten-year-old CO sensors, and were once used in electromechanical timers that controlled dual-rate electricity meters. When photography was analogue such 3V sources powered the film advance motors and other electronics of simple cameras.
  • #10 21153936
    error105
    Level 14  
    What could be the real delays in open space and 500m distance? Do you have any idea how to estimate such a thing to know exactly how much the transmission was?
  • #11 21154014
    TvWidget
    Level 38  
    error105 wrote:
    What could be the real delays in open space and 500m distance?
    .
    Signal propagation over this distance is approximately 16 nS.
  • #12 21154142
    TechEkspert
    Editor
    Packet transmission is quite slow at high SF values, so the method of modulation will affect the time from preparation of the data for sending, to the appearance of the data in the receiver buffer. This can be as low as 500ms for 100B and more for slower transmissions: How much range do LoRa E32 modems have? .
  • #14 21155000
    dktr
    Level 25  
    >>21154858 If I was transmitting something from where I have these receivers - on that high object then maybe I would be interfering with something. But there is only a receiver there. The transmitter is currently hidden 6.5 km away from the city in the middle of a forest and sends temperature measurements every 2 minutes, and the sending itself takes 0.4s so very little chance of me interfering with anything. In addition, from the graph of the background level from the receivers on that frequency, you can see that nothing is transmitting there, and if it were transmitting it would sooner interfere with my correct reception from that sensor - and as of now 100% of the readings are correctly received. It's a small town, and not much industry here.
  • ADVERTISEMENT
  • #15 21155067
    TvWidget
    Level 38  
    LoRa is a wideband modulation that is immune to interference. However, a signal with such modulation increases the noise floor on many channels. For this reason, special regulations for the use of such signals regarding output power and bandwidth occupation factor were added a few years ago. If I remember correctly, the least restrictive requirement specifies that for the 868 MHz band the interval must be at least 1000 times longer than the transmit time. For 433 MHz the restrictions may be greater.
  • #16 21163571
    maras77
    Level 21  
    dktr wrote:
    .
    Battery from a disposable e-cigarette with an electronic circuit on a wooden background. .
    .

    A little off-topic:

    I see a link from a disposable e-cigarette ;)

    I once spotted a commercial, with used disposable e-cigarettes in the electro-junk in a ladybird shop and took them in. As a result, I have a dozen or so of these cells, which are actually very good quality.

    Moderated By ArturAVS:

    3.1.14. Posting that is inconsistent with the topic of the relevant forum section or discussion thread.

    .
  • #17 21171035
    jarekgol
    Level 39  
    @dktr what do the red and blue circles on the map mean?
    What kind of map are you using? I used to train geoportal for a similar game.

    @error105 I don't know exactly how much you need, but you could try equipping both devices with gps, synchronise the time and try to work something out from that. Obviously if the times are of the order of ns then that won't work.
    The question is also from which point in time to measure them, is it from the call of the send function, or from the end of the serial transmission to the chip, or something else?
  • #18 21171049
    dktr
    Level 25  
    >>21171035 The colour of the dots indicates which receiver received the signal, blue dot both receivers, red dot one receiver. I always use openstreetmap because it has a nice api and you can se locally host everything including map files.
  • #19 21171107
    acctr
    Level 38  
    maras77 wrote:
    Thanks to this, I have a dozen or so of these cells that are really very good quality.

    What are these cells and what do you charge them with?
    Helpful post? Buy me a coffee.
  • ADVERTISEMENT
  • #20 21206387
    maras77
    Level 21  
    acctr wrote:
    maras77 wrote:
    Thanks to this, I have a dozen or so of these cells that are really very good quality.

    What are these cells and what do you charge them with.
    .

    Cells, like cells. They have a high discharge current. I don't know about the durability. And I charge them with a CC/CV power supply.
  • #21 21288244
    mariusztol64
    Level 1  
    Tell me if you would be able to make an interface to the reticulum network?
    https://reticulum.network/manual/interfaces.html

    PIPE is the simplest way. All you need to do is write a program , which sends everything it gets on the input into the ether and everything it gets from outside it sends to the output. Of course KISS is the best because it has more options and you can use more programs. But reticulum already has ssh so I don't know if that wouldn't be enough.

    Two more technical questions.
    * How about power supply at -20 degrees?
    * Is there any chance of power from solar panels?

    Isn't it better to just use the esp8266 alone? After all, wifi also works at such distances and greater, so why multiply modules? Besides, if you have connectivity on lora, you don't connect to the internet. And it would be good for the node to combine wifi, lora 868 and 443MHz
  • #22 21616239
    dktr
    Level 25  
    Update.

    I have currently measured several lakes in the Ostróda area.
    Here you can see the results of the measurements http://meteo.lipowa.net/lora_map.php
    The device that sends the measurements is STM32G030F6P6 + SX1278 and two DS18b20 pieces, all powered directly from a 3.6 V 17 Ah lithium battery without any stabilisers along the way. I have previously used an ESP8266 in such a configuration and was not able to go below 25 μA in sleep mode, then I tested an RP2040 and it was a total failure as it was something around 100 μA, an ESP32C3 9 μA, and this STM 3.9 μA, I still have the WCH and nRF microcontrollers to experiment with, but that will be in my own time. The sensor sends a measurement every 300 seconds or so, the STM has an internal stabiliser which it uses to measure the voltage the microcontroller is powered by, it works surprisingly accurately in the 3.6-3.0V range I'm interested in.

    PCB with STM32G030F6P6 microcontroller, SX1278 module, and two DS18b20 sensors powered by a lithium battery.
  • #23 21616369
    TechEkspert
    Editor
    Interesting results, how did you measure the current draw of the new design?
  • #24 21616839
    metalMANiu
    Level 21  
    >>21616239 .
    Great project.
    What realistic battery life do you anticipate?
    How have you protected the equipment from the weather and theft? Theoretically from the map you have linked it is possible to locate the transmitters.
    Have you calibrated the sensors to discard those that give outlier results?
    I ask because the water in Lake Sajmino seems to be very warm.
  • #26 21616867
    dktr
    Level 25  
    >>21616840 I doubt it. I did this to satisfy my curiosity and see the annual temperature graph ;) .

    UNI-T UT70B multimeter on the ground displaying a measurement; test lead connected.

    >>21616839 Theoretically, it should last about nine years. The unit itself is in an airtight box and any holes are only on the underside - the antenna, the cable to the sensor and the sensor itself. On the map is what is intentionally not exact - strongly approximated, although a little patience - a measurement every 300 seconds a few seconds, a trip with the radio and you can locate, the only ones who move there are fishermen, I hope no one notices, it hangs quite high on a tree and the cable is hidden. The sensor is a DS18B20 I use dozens of them, I even calibrated some in an accredited lab, from what I remember they had an error of about +0.3°C at 20°C and -0.1°C at 0°C. In Lake Sajmino I think someone fished the sensor out and it's floating by the reeds, I need to go there and see what happened.

    >>21616369 With a UNIT UT-70B meter
  • #27 21618209
    jarekgol
    Level 39  
    Which map did you use to show the "balloons" ?
    Ok, from what I see OpenStreetMap and Leaflet, what are your impressions of working with such a set?
    I played around with our geoportal and had to guess a lot of things.
  • #28 21618277
    acctr
    Level 38  
    dktr wrote:
    I did this to satisfy my curiosity and see the annual temperature graph
    .
    Environmental organisations who actually care about caring for the ecosystem rather than screaming and sticking to the tarmac should take an interest.
    Helpful post? Buy me a coffee.
  • #29 21618511
    dktr
    Level 25  
    >>21618209 Openstreetmap is very convenient, there was no problem, I always use it for such purposes.
    >>21618277 Only that this would already have to be standardised somehow - measurement at a specific depth etc. Currently I am trying to place the sensor so that it is near the bottom at a depth of 1m. Today I added Lake Pausenka, where there are a lot of reeds near the shore and it is generally shallow, so the sensor is on as much as 30m of cable in a small glass jar flooded "under the cork" with hot glue. There are still Morliny and Szeląg lakes left, I think there will be no problem with the range even though it is over 8km, currently the furthest sensor is in Górka - 6.5km and the reception efficiency is 100%. I am still planning to make a "measuring tube", i.e. a few dallas one under the other at a fixed distance so as to have a temperature measurement at a specific depth, the whole thing encased in resin to make it airtight. In the photo, the device is attached to a tree near Lake Pause.


    A sensor device in a glass jar attached to a tree trunk by Lake Pauzeńskie. .
  • #30 21618798
    jarekgol
    Level 39  
    dktr wrote:
    making a "measuring tube" i.e. several dallas one under the other at a fixed distance so as to have a temperature measurement at a specific depth,
    .
    I've always wanted to do something like this and observe the stratification of the water, just in some small tank, not a lake. And observe the temperature in a high flat.
    Assuming it's going to be one set up, then you could probably use ordinary NTCs.
📢 Listen (AI):

Topic summary

The discussion revolves around experiments with LoRa technology using SX1276 radio modules in the 70 cm band. The author built multiple devices, including two gateways and several transmitters, to evaluate the range and capabilities of these modules. Key points include the importance of bandwidth occupancy for long-range transmission, the use of lithium batteries for power efficiency, and the challenges of signal transmission delays. Participants also discussed the modulation methods, potential interference with other devices, and the use of GPS for synchronization in experiments. The author shared insights on battery performance in low temperatures and the advantages of using non-rechargeable batteries over rechargeable ones.
Summary generated by the language model.
ADVERTISEMENT