Author Topic: LoRa-RFM96 433MHZ range test -->5.2 miles, 8.3km !!!  (Read 15621 times)

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
LoRa-RFM96 433MHZ range test -->5.2 miles, 8.3km !!!
« on: February 19, 2016, 10:12:46 PM »
  Hello all,
I am in the early stages doing some range test with the LoRa modules RFM96 (433MHZ)  I just got an amazing result of 2.2 miles..  TX and RX faces a lot building in the way.. my transmitter antenna is in inside the balcony   in my apartment at first floor, antenna is at  2 meter high (1/4 lambda), the receiver is  setup with  3dbi antenna in my car, I am pretty sure that the range at open field  will be much better,  I will provide more information if anyone is interested.  for my test I use the radiohead library the example under the RFM95 named as  rf95_reliable_datagram_client,  to get that range I select this  setModemConfig(Bw31_25Cr48Sf512);  in  the RH_RF95.cpp. I also change the power output to 20dbm -->  driver.setTxPower(20, false);
Now I am at the point where i need some basic help on the best way to send  usefully data for telemetry purposes. I will post  my question  related to that matter on  different thread
« Last Edit: April 15, 2016, 07:21:11 PM by luisgcu »
..Knowing other is wisdom,
knowing yourself is Enlightenment..

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #1 on: February 19, 2016, 11:30:43 PM »
I couldn't   wait until tomorrow so I  did a second test getting the "hello message" sent  from tx at home  at a decent  distance of 2.95 miles (4.7km). :)
..Knowing other is wisdom,
knowing yourself is Enlightenment..

WhiteHare

  • Hero Member
  • *****
  • Posts: 1298
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #2 on: February 19, 2016, 11:43:38 PM »
That's very impressive!  How do the relative elevations compare?  I'm guessing it's either line-of-sight or something close to that?

I live in a hilly area, and I was getting great range up until too much hill got between the Tx and Rx.  It did better than LOS, though.  It was picking up a lot of signals bounced off buildings that it couldn't have gotten if it were purely LOS. 

When the signal "absolutely, positively has to get there," LoRa is a great choice, even for shorter distances like around the house.  The incremental cost is reasonably low considering what you get, and it comes with a top notch FSK radio similar to the RFM69 in case you want to run at higher bitrates.  Because of the higher Tx time, though, for battery applications I tend to think of LoRa as an awesome fallback option in case the primary FSK communications aren't getting through.

« Last Edit: February 20, 2016, 09:29:14 AM by WhiteHare »

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #3 on: February 20, 2016, 09:19:28 AM »
I would  say that is not true LOS what I have  between  TX and RX, maybe its  lot of reflection on buildings this a picture of how my TX is installed, I put inside that plastic pipe to protect moteino from rain.
..Knowing other is wisdom,
knowing yourself is Enlightenment..

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #4 on: February 20, 2016, 10:33:19 AM »
This is a new test while  I drive to Miami  I got signal 5.2 miles from  home.  I just couldn't guess  how far the range maybe if  I can place the antenna  on the building roof.
terrain is pretty flat but there are a lot building and bridges in the LOS.
« Last Edit: February 27, 2016, 05:39:52 PM by luisgcu »
..Knowing other is wisdom,
knowing yourself is Enlightenment..

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5959
  • Country: us
    • LowPowerLab
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #5 on: February 20, 2016, 08:51:54 PM »
Quite amazing, many thanks for sharing this!
BTW I have reports from another user who has gotten a link of over 160km with RFM69 (not LORA!).
Yes that is one hundred sixty kilometers.
The catch and essential details is that he deployed a weather balloon, the bitrate was of course the lowest possible.
So at high elevation, the link can be over 200km even with FSK radios.
Maybe I will share the info and code in the blog or forum when I catch up with the backlog.

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km)
« Reply #6 on: February 27, 2016, 05:06:09 PM »
What you descrive is very impressive  result.  I just tested my rfm69 , no tweaking at all and  range  is very short. . I still want to do more test  and see what is the magic to get longer ranges distances.

I see that the rfm69 library have lot of options for the radio mode need to explore which  is the optimal  to achieve longer ranges.
I know the lower the frequency ,then rf signal can easily go through walls and other objects in the  Los , that's why I like the 433mhz but the problems  that there is some restrictions  using that frecuency in the US.
The photo below is my about my moteino  with the rfm96 where ever I go I get that with me to see if I can catch the signal from the server.  Love that little  LoRa module
« Last Edit: February 28, 2016, 10:43:53 AM by luisgcu »
..Knowing other is wisdom,
knowing yourself is Enlightenment..

dave_sausages

  • Newbie
  • *
  • Posts: 49
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #7 on: March 13, 2016, 07:12:22 PM »
@Luisgcu

Is it possible for you to post your code for the settings you changed? I'm trying to increase the range of my Moteino LoRa, and I'm struggling to 1) decide what settings to use, and 2) how to change them.

Hopefully your code will be a stepping stone for me :)

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #8 on: March 14, 2016, 08:41:25 AM »
Happy to share what I got, which is not much.I basically use the example under rf95 I tried with  rfm95_client and server  and also tried rf95_reliable_datagram_client  and server,  these 2 example work  perfectly without any modification.   If you want to get long range you need to change few parameter inside the Radio head library.  I notice that lowpowerlab  has the same radio head library  with some updates  but I am no using  that library  I have to investigate  if it has some update  or improvements  for the RFM95.
By the way I  am using the rfm96  LoRa  433mhz  not the RFM95  which  is 915mhz.
When I get home today I'll share more details.
Regards
..Knowing other is wisdom,
knowing yourself is Enlightenment..

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #9 on: March 14, 2016, 09:39:20 PM »
Happy to share what I got, which is not much.I basically use the example under rf95 I tried with  rfm95_client and server  and also tried rf95_reliable_datagram_client  and server,  these 2 example work  perfectly without any modification.   If you want to get long range you need to change few parameter inside the Radio head library.  I notice that lowpowerlab  has the same radio head library  with some updates  but I am no using  that library  I have to investigate  if it has some update  or improvements  for the RFM95.
By the way I  am using the rfm96  LoRa  433mhz  not the RFM95  which  is 915mhz.
When I get home today I'll share more details.
Regards

here is the example that I am using.. is a  little messy but it works copy the the content of the zip file  inside the radiohead library.
« Last Edit: March 15, 2016, 07:49:39 PM by luisgcu »
..Knowing other is wisdom,
knowing yourself is Enlightenment..

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #10 on: March 20, 2016, 10:42:31 PM »
I added a GPS to one Moteino LoRa, that allow  me to  send the position from client to server, every time server get new position it is stored in a SD card for further  analysis or upload to google earth.
I am trying to understand the result of the different possible combination of the RFM96 config registers.
..Knowing other is wisdom,
knowing yourself is Enlightenment..

WhiteHare

  • Hero Member
  • *****
  • Posts: 1298
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #11 on: March 21, 2016, 12:39:07 AM »
Nice.  Which GPS and GPS library do you like the best?

If you were to load and use Semtech's LoRa calculator, you can what-if different permutations of register settings and see what effect it has on the link budget.  It's one way to develop a gut feel for what's possible.

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #12 on: March 21, 2016, 08:53:15 AM »
Thanks Whitehare , the LoRa modem  calculator  seems  to be interesting, I will play with it .
I  am using   (ardupilot Ublox  gps) I bought that from ebay just for this test. The library I am using is the  TinyGps++.
On the receiver side I  am using an additional arduino connected  to the moteino via I2C.  This arduino have the SD card ,I was unable to get working properly  the Rfm96 and sd card sharing the same SPI.
I will share all that project when I cleanup the messy code and put some comments  on it.
« Last Edit: March 21, 2016, 08:55:18 AM by luisgcu »
..Knowing other is wisdom,
knowing yourself is Enlightenment..

WhiteHare

  • Hero Member
  • *****
  • Posts: 1298
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #13 on: March 22, 2016, 01:35:33 PM »
Sounds reasonable.  If you ever do decide to take another run at directly connecting the SD card, you may want to give Arduino IDE 1.0.6 a try.  It may be a longshot, but at least on the Moteino R4 reverting to 1.0.6  resolved some bizarre SPI issues I was having wrt the Winbond SPI Flash memory. 
« Last Edit: March 22, 2016, 09:44:31 PM by WhiteHare »

luisgcu

  • Newbie
  • *
  • Posts: 29
  • Country: us
Re: RFM96 range test 2.2 miles ( 3.7km) -->..Update-->5.2 miles !!!
« Reply #14 on: March 22, 2016, 10:21:13 PM »
that tip you mention related to the arduino ide is interesting... I read a lot a bout SPI problems in regard the RFM69 , I  tried different things but none get the problem resolved.
I have another question and is related to the payload that I sending  basically I am sending this :

struct Gpsdata{
 double latitude;
 double longitude;
 //long data;
  };
Gpsdata MyGpsdata;
uint8_t*GpsPayload;

void setup{
 
GpsPayload = (uint8_t*)(&MyGpsdata);

}

loop{

if.... condition{
 if (manager.sendtoWait(GpsPayload, sizeof(MyGpsdata), SERVER_ADDRESS))
       {

}

the doubt that I have now in regard to  my program I am not  sure if  I am setting the message length  therefore I tend to  believe I am sending the RH_RF95_MAX_MESSAGE_LEN without need... in my struct i only  have 16 byte, so I am not sure if I am sending only those 16 byte or the whole 255 octets -header..  the question is if I have to set the exact number of octets?
or automatically the Sendtowait  do the calculation of how many bytes  using the "sizeof(MyGpsdata) ??




..Knowing other is wisdom,
knowing yourself is Enlightenment..