Lora: Documentation vs. Example code regards OTAA

[robert-hh]

There is a discrepancy between the documenattion and the example cod ein the lib regarding the auth paramter.
The documentation tells, it's a 2-tuple of (app_eui, app_key)
while the example code uses a 3-tuple, of (dev_eui, app_eui, app_key).
None of them works for me, but besides that, what's the right one?

[asavvin]

Hello!
I am having issue with activating using method otaa lopy with ttn gateway fot US915.
whats correct code i should use I searched a lot of topics over weekend and could not find correct code.
Thank you

Same time using ttn node US915 with ttn gateway activation worked as shown.

[robert-hh]

@ssmith Besides that, no luck. Simple device-to-device works, the power getting out of the antenna looks similar, with the FiPy a little bit (25% = 2 dB) stronger. But that may be within the setup-variation.

[robert-hh]

@ssmith Yes, I'm using TTN. As far as I can could read, the limit is for downlink messages, from TTN to the node. Uplink, to TTN, I did not experience any limit.

[ssmith]

@jmarcelino Wasn't able to get EU868 to work with LoRaWAN. Gave me a frequency error. I was able to set it up and send raw LoRa between 2 LoPy4's at both 868 and 915 bands without changing the FW. Looks like it is in the LoRaWAN stack.

[ssmith]

@robert-hh Are you going to TTN? I've sent thousands of messages there in a day during testing.

[robert-hh]

@jmarcelino Cannot try more, I think I expired my downlink message quota (of 10?) for the day.

[ssmith]

@jmarcelino I also tried AU915 with the same result. I'll try EU next. I assume with the generic FW now that it should accept it.

[ssmith]

@jmarcelino It's there, that was a different config file from my previous test just to show the frequency. Here's the config from the nano_gateway file.

        self.lora = LoRa(
            mode=LoRa.LORA,
            region=LoRa.US915,
            frequency=self.frequency,
            bandwidth=self.bw,
            sf=self.sf,
            preamble=8,
            coding_rate=LoRa.CODING_4_5,
            tx_iq=True
        )

[jmarcelino]

@ssmith
Could you add region=LoRa.US915 as a parameter in both lora.init() please?

[ssmith]

@jmarcelino, region=LoRa.US915,. This worked fine with 1.15. and here is lora config

lora.init(mode=LoRa.LORA, frequency=918900000, power_mode=LoRa.ALWAYS_ON, tx_power=20, bandwidth=LoRa.BW_125KHZ, sf=7,
              preamble=8, coding_rate=LoRa.CODING_4_8, tx_iq=False, rx_iq=False)

[jmarcelino]

@ssmith
Which region are you in please and what frequency is LORA_FREQUENCY set to in config.py?

This error seems related to the recent LoRa region changes in 1.16 but it's strange that the nanogateway is able to listen on the correct frequency but not transmit on 923300000.

[ssmith]

@jmarcelino Since updating to 1.16 I'm getting this error when trying to connect.

[   388.642] Received packet: {"rxpk": [{"data": "AFlERE/vvq3e6FJQ/v+krjC7ifKetjw=", "time": "2018-02-19T13:38:40.120571Z", "chan": 0, "tmst": 388895160, "stat": 1, "modu": "LORA", "lsnr": 5.0, "rssi": -17, "rfch": 0, "codr": "4/5", "freq": 918.9, "datr": "SF7BW125", "size": 23}]}
[   388.675] Push ack
[   393.694] Pull rsp
Unhandled exception in callback handler
Traceback (most recent call last):
  File "lorawan_nanogateway.py", line 386, in <lambda>
  File "lorawan_nanogateway.py", line 345, in _send_down_link
ValueError: frequency 923300000 out of range

[robert-hh]

@jmarcelino Agreed, because that's what is see with ABP mode too.

[jmarcelino]

I think the problem you're having with OTAA is related to you not receiving downlinks properly https://forum.pycom.io/topic/1254/ttn-otaa-join-acknowledgement-not-being-received/13 and not the Dev EUI

[robert-hh]

@jmarcelino In my tests so far, neither using the 2-tuple or the 3-tuple made OTAA succeed. But I'm still trying more this night.

[jmarcelino]

Hi @robert-hh
Do you mean if you use auth=(dev_eui, app_eui, app_key) with OTAA you don't see the Device EUI change to the one you are specifying?

[robert-hh]

@seb No change when running the code in either way.

[seb]

I agree, I am now updating the example in the libraries repo. What do you mean it makes no difference? Do you not see the dev_eui change, or are you having more general LoRa issue?

[robert-hh]

@seb Thanks. As said, it makes no difference, and I did not take the time to look into the code. But maybe it could be harmonized.