RAK3172 Insufficient Coverage?

Hi everybody,

We have been working on communicating two RAK3172 modules in P2P mode for a while.

At first we worked with 868 MHz and after 300-400 meters communication between modules was lost.

Then we decided to set the frequency to 433 MHz. We connected 433 MHz antennas with 5 dBi power to our RAK3172 modules. And then in the final test we did today, we were able to reach 400 meters in the city (the RSSI was around -115 and -120 at that moment). We couldn’t get any data after 400 meters.

The transmitter (TX) RAK3172 from the modules was located on the terrace on the top floor of the building where I was located. The receiver (RX) module was located in a car 400 meters away. This was the maximum range we were able to reach during testing.

The parameters we applied in the communication test:

double myFreq = 433000000;
uint16_t sf = 12, bw = 0, cr = 0, preamble = 8, txPower = 22;

I can’t understand why the distance is so low. Are we doing something wrong, or can we communicate with the RAK3172 5 dBi antenna and the above parameters for a maximum of 400 meters?

Hi @whydont

The RF path inside the RAK3172 is matching for either 800-920 MHz (RAK3172-H) or ~400MHz (RAK3172-L). When using our RUI3 firmware, you cannot set a frequency lower than 800 MHz on the RAK3172-H module. What firmware are you using? If you are using a custom firmware created with STM32CubeIDE, are you adjusting the LoRa part as described in our Low-Level-Development tutorial?

If not, the STM32WL5E will not work as expected, which could explain the short range.

1 Like

I’m using latest firmware and editing code at Arduino IDE. For setting low freq. mode, i pulled down PB12 and set frequency with code and with AT+PFREQ=433000000 command.

So what’s the problem?

For 433MHz, even you pull down the pin, the RF path is still not matching for 433MHz, it is still for 800-920MHz.

What is between the two nodes when you do the tests? You said you test in the city? Are there buildings or other structures between the two nodes?

Why do you say the RF path is not matched for 433 MHz? The module I have, does not say RAK3172(L) or RAK3172(H). What would prevent me from using it for 433 MHz?

Yes, I tested it in the city. Although I do not insert a building directly between the two modules, buildings, trees, etc. within 400 meters. exists.

That’s how my module looks like (No “H” or “L” indicator):

Then it is a -H version.

If you look into antenna matching, you will learn that the R/C/L matching between the antenna and the RF pin of the STM32WLE5 has to match with the frequencies it should work with the best efficiency.

The values of all marked components are different between the H and the L version.

Of course you can always fake a different module, but don’t complain if the performance is not as expected.

You did not answer my question regarding the test environment.

How did you decide it was a -H version? I can’t understand because I can’t see H or L on the module.

I explained the test environment in my previous answer.

Test was on city. There are buildings (not fully blocking), trees, etc. between nodes.

It is really confusing. Even if we say my module is only for 868 MHz, isn’t 400 meters insufficient?

Yes, the LoRa transceiver can be set to 150Mhz to 960MHz. But that is INSIDE the chip. It still needs a RF path matching and correct frequency antennas to get a usable range.

I ask one more time.

What is the test environment?

When I go out with a mapper from my house, I get signal until the end of my street, as soon as other buildings are between the node and the receiver, I loose the signal. In some spots with less structures between the two devices I can get 700m, in some spots I get no connection at all.

This is a heat map when I tested the device. As you can see there are many areas where the device could not reach the gateway.

1-) How did you decide that my RAK3172 module is a -H version, even though there is no “H” or “L” phrase on it as it says in other modules?

2-) If in the features section (150- 960 MHz) is written, is it an empty expectation to be able to communicate at the frequency I want in this frequency range without saying “fake etc.”?

No -H on the label means H version.
-L on the label means L version.
(I) on the label means it has IPEX connector
no (I) on the label means it has no IPEX connector.

City environment can mean anything. The picture makes it clear. If there is direct line of sight, the range is too less. I get more with our small PCB antennas that we use for evaluation.

The 5dBi antenna you are using is our 5dBi antenna for 863-870 MHz? You should getter better range with it.

With LoRa P2P you can test with different SF and bandwidths to increase the range. There is some information in this article

Thanks. There is no -H or -L in my RAK3172 module. There is only the (I) label.

I can’t say the view is direct line of sight. But i can say it is semi-direct sight.

No, it is not your antenna. It is from another brand (Probrably made in PRC).

I set the SF parameter as SF12 which provides max. range. Also bandwith set to 0 (125) for max. range.

The antenna which i used:

For what frequency is that antenna? Is it for EU868 frequencies?

Yes. When I worked with 868 MHz, I used 868MHz of this antenna. When I worked with 433 MHz, I used a similar 433 MHz antenna.

Can you explain this a little more?

Some stuff to read:
What are the Factors Affecting LoRaWAN Range
What is the real range of LoRa?

Doesn’t matter if it is about LoRaWAN, the transmission method is the same with LoRa P2P

So you are saying that 400 meters is the maximum range I can get, unless I position the antennas as indicated in the image below (ie, as long as there are buildings etc. in between). Do I understand correctly?

Buy a proper antenna as it’s very important and most of “Alixexpress” ones are quite terrible and not for frequencies indicated on it !
I get few kilometers range easily with RAK3172 and a proper 2.3dB antenna (one from RakWireless with proper cable too).


Can you see the second unit from the first? This is important due to the physics of radio waves. “not fully blocking” is vague.

1 Like

Influential. What is your test environment? Is there a direct line of sight between the receiver and transmitter modules? Or are there buildings in between?