LoPy4: operation problem US_915

  • I'm using a LoPy4 in LoRaWan mode and I have to configure the US_915 channels like this:

    CH2 902.7 DR0-DR3
    CH3 902.9 DR0-DR3
    CH4 903.1 DR0-DR3
    CH5 903.3 DR0-DR3
    CH10 904.3 DR0-DR3
    CH11 904.5 DR0-DR3
    CH12 904.7 DR0-DR3
    CH13 904.9 DR0-DR3
    CH64 903.0 DR4

    the network server manages this configuration, the channels that are not present in the list are not managed. Below is the code executed for the configuration of the channels:
    the data rate varies from 1 to 3 on the first 8 channels by necessity, the payload is too heavy for dr 0 (16byte)

    for i in range(0, 71):
    lora.add_channel(index=2, frequency=902700000, dr_min=1, dr_max=3)
    lora.add_channel(index=3, frequency=902900000, dr_min=1, dr_max=3) 
    lora.add_channel(index=4, frequency=903100000, dr_min=1, dr_max=3)
    lora.add_channel(index=5, frequency=903300000, dr_min=1, dr_max=3)
    lora.add_channel(index=10, frequency=904300000, dr_min=1, dr_max=3)
    lora.add_channel(index=11, frequency=904500000, dr_min=1, dr_max=3)
    lora.add_channel(index=12, frequency=904700000, dr_min=1, dr_max=3)
    lora.add_channel(index=13, frequency=904900000, dr_min=1, dr_max=3)
    lora.add_channel(index=64, frequency=903000000, dr_min=4, dr_max=4)

    I have encountered the following problems:

    . RF signal disturbed on 50% of transmissions (controlled with spectrum analyzer)
    . about 60% loss of packets sent
    . very different RSSI and SNR values between a message and the next one
    . firmware lock in 10% of cases on the package send function
    . the transmission channels of the lopy are sometimes different from those set by the code

    Do you have any advice to give me?

  • @_grunf Did you figure it out?

  • @robert-hh
    Ok, in the next few days I will publish the results of the new tests on this topic.

  • @_grunf With that gateway and properly installed antennas you should not have many variations. The channel set-up looks fine to me, but I never tried to devices with the US_915 settings. That could cause some trouble here.

  • @robert-hh
    the gateway is this:
    Next week I will do more accurate tests with the spectrum analyzer following your advice.
    According to you, is the channel creation procedure correct?

  • @_grunf OK.

    • everynet is the network, but are you using your own gateway or just one which is by chance in reach? If yes, then both the varying rssi levels and the packet loss are to be expected.
    • I do not assume that the RF signal is disturbed. What you see is most likely cause by the way you run the spektrum analyzer. To see a clean peak of your transmitted messages, the spektrum analyzer's frequency sweep has to be synchronized with the message transmissions, since the RF signal is present only for the time of transmission.

  • Hi @robert-hh ,
    thanks for the quick reply

    which gateway are you using? And is it the same all the time?
    always the same gateway:

    how do you say the RF signal is disturbed?
    sometimes during the sending of the message I see a clean peak of signal on the desired frequency, sometimes on the whole band there are so many spurious frequencies at the moment of transmission

    do you wait long enough between packages?
    about 2 minutes

    what do you call very different than RSSI and SNR
    RSSI: from -120db to -30
    SNR: from -10 to + 12

  • @_grunf Just a few questions:

    • which gateway are you using? And is it the same all the time?
    • how do you tell that the RF signal is disturbed?
    • do you wait long enough between to packets?
    • what do you call very different with respect to RSSI and SNR

    Please be aware that the Lora server can change the channel frequencies, and the LoRa stack supports that.

Pycom on Twitter