Gateway sends join-acept but device activateOTAA() returns -1116 anyway

[quintesse]

Given the title what could I be doing wrong?

It seems to me that the RadioLib configuration should be at least correct enough to send the join request and get it accepted. I also haven't noticed any packet loss (device and gateway are in the same room) so I'm assuming it's not that the device isn't receiving the reply.

The code I'm using is functionally the same as LoRaWAN_Starter.

The Live Data in the console shows "js.join.accept", followed by "ns.up.join.process" and finally a "as.up.join.forward".

I haven't added any logs but can provide them.

[StevenCellist]

Please provide your hardware setup, software setup and the (debug protocol) log of a full join cycle, with millisecond timestamps enabled in your terminal.

[quintesse]

The hardware is a XIAO ESP32S3 with a Wio-SX1262 LoRa module.
Not sure what you want to know about the software setup, a link to a code repo?

And I have some logs but they are without timestamps, sorry, I'd have to figure out how to configure that 😅

12:40:04.738 > INF: Initialise the LoRaWan radio...
12:40:04.788 > INF: Joining the LoRaWAN Network...
12:40:04.788 > RLB_PRO: Setting up dynamic channels
12:40:04.788 > RLB_PRO: UL:   0 1 868.100 (0 - 5) | DL:   0 1 868.100 (0 - 5)
12:40:04.788 > RLB_PRO: UL:   1 1 868.300 (0 - 5) | DL:   1 1 868.300 (0 - 5)
12:40:04.789 > RLB_PRO: UL:   2 1 868.500 (0 - 5) | DL:   2 1 868.500 (0 - 5)
12:40:04.789 > RLB_PRO: [MAC] 0x03
12:40:04.789 > RLB_PRO: 00000000: 20
12:40:04.789 > RLB_PRO: LinkAdrReq: dataRate = 2, txSteps = 0, nbTrans = 0
12:40:04.790 > RLB_PRO: LinkAdrAns: 07
12:40:04.790 > RLB_PRO: [MAC] 0x04
12:40:04.790 > RLB_PRO: 00000000: 07                                               .
12:40:04.790 > RLB_PRO: DutyCycleReq: max duty cycle = 1/2^7
12:40:04.790 > RLB_PRO: [MAC] 0x05
12:40:04.790 > RLB_PRO: 00000000: 00 d2 ad 84                                      ....
12:40:04.790 > RLB_PRO: RXParamSetupReq: Rx1DrOffset = 0, rx2DataRate = 0, freq = 869.525
12:40:04.797 > RLB_PRO: [MAC] 0x08
12:40:04.797 > RLB_PRO: 00000000: 01                                               .
12:40:04.797 > RLB_PRO: RXTimingSetupReq: delay = 1 sec
12:40:04.797 > RLB_PRO: [MAC] 0x09
12:40:04.797 > RLB_PRO: 00000000: 05                                               .
12:40:04.797 > RLB_PRO: [MAC] 0x0c
12:40:04.797 > RLB_PRO: 00000000: 65                                               e
12:40:04.797 > RLB_PRO: ADRParamSetupReq: limitExp = 6, delayExp = 5
12:40:04.797 > RLB_PRO: [MAC] 0x0f
12:40:04.797 > RLB_PRO: 00000000: fa                                               .
12:40:04.797 > RLB_PRO: RejoinParamSetupReq: maxTime = 15, maxCount = 10
12:40:04.867 > RLB_PRO: 
12:40:04.867 > RLB_PRO: PHY:  Frequency = 868.300 MHz, TX = 16 dBm
12:40:04.868 > RLB_PRO: LoRa: SF = 10, BW = 125.0 kHz, CR = 4/5, IQ: U
12:40:05.246 > RLB_PRO: JoinRequest sent (DevNonce = 0) <-- Rx Delay start
12:40:05.247 > RLB_PRO: 00000000: 00 00 00 00 00 00 00 00 00 06 d0 93 c0 7e d5 b3  .............~..
12:40:05.247 > RLB_PRO: 00000010: 70 00 00 87 ca 84 cf                             p......
12:40:05.248 > RLB_PRO:
12:40:05.248 > RLB_PRO: PHY:  Frequency = 868.300 MHz, TX = 16 dBm
12:40:05.249 > RLB_PRO: LoRa: SF = 10, BW = 125.0 kHz, CR = 4/5, IQ: D
12:40:10.243 > RLB_PRO: Opening Rx1 window (231 ms timeout)... <-- Rx Delay end 
12:40:10.474 > RLB_PRO: Closing Rx1 window
12:40:10.475 > RLB_PRO:
12:40:10.475 > RLB_PRO: PHY:  Frequency = 869.525 MHz, TX = 16 dBm
12:40:10.476 > RLB_PRO: LoRa: SF = 12, BW = 125.0 kHz, CR = 4/5, IQ: D
12:40:11.243 > RLB_PRO: Opening Rx2 window (688 ms timeout)... <-- Rx Delay end 
12:40:11.931 > RLB_PRO: Closing Rx2 window
12:40:11.931 > ERR: Failed to join LoRaWan network: #-1116

[StevenCellist]

Not sure what you want to know about the software setup, a link to a code repo?

Preferably, a dump here of the first part of config.h (redact AppKey/NwkKey, the EUIs can be intercepted anyway) up until LoRaWANNode node... as well as your main file up until node.activateOTAA(). If you use ArduinoIDE, it comes with a builtin button for timestamps; if you use PlatformIO, you can add timestamps using monitor_filters = time in your .ini. Otherwise, check documentation of your terminal program.

[quintesse]

I've updated the log with a new one with timings and the code can be found here: https://github.com/quintesse/arduino/blob/main/ESP32S3/tof_oled_lorawan/src/main.cpp

[HeadBoffin]

The code I'm using is functionally the same as LoRaWAN_Starter.

The devil being in the details, perhaps share that or use LW Starter.

device and gateway are in the same room

Tiny device says "Knock knock", giant responds in booming voice "You're in!", device totally overwhelmed with the volume & doesn't understand anything. LoRa = Long Range - 5m & a brick wall separation is a typical minimum - if your JR RSSI is -50 or above it will need moving.

TL;DR: You've got known good hardware, the MCU & radio are talking, so this looks likes the classic "too close" problem.

[HeadBoffin]

But because this is adding to my detailed files for training notes, I'll be a little kinder. But only a little.

Once upon a time an old fa&t wrote the notes.md and it was deemed good because it had all the main gotchas detailed in it. Including the DevNonce issue when developing a device ...

[quintesse]

In respect to that old fa®t I had already read that doc from beginning to end but that doc itself doesn't talk about DevNonces, right? And in the troubleshooting section of the Wiki it points to the only thing it says about result code -1116 is "Not all uplinks to LoRaWAN server result in a downlink. This is not a problem, but rather a common behavior in LoRaWAN", which sounds like "don't worry about it" but then all the code examples treat it like a failure, so that confuses me. But I'll stop pestering you with questions... at least until I've RTFM.

[HeadBoffin]

RadioLib/examples/LoRaWAN/LoRaWAN_Starter/notes.md

Line 61 in 577a076

When many sensors are big deployed, a device is registered, batteries put in, it joins and gets on with sending data for the next few years. For development purposes we need to turn off one of the security settings so that you can join & uplink out of the normal sequence that a device in the field would do.

[HeadBoffin]

Once we've got you going, we'll have to take this and figure out the messages - every JR should result in a downlink.

For clarity, DevNonce is the technical name for "the join nonce"

[quintesse]

Just a small update. After having installed a new antenna on my device everything seems to working much better. The RSSI went up from -130 to -67 (without having moved it or the gateway) so that's good too.

I do get some failed sends at times, but they actually arrive at the gateway, so perhaps the "ack" gets lost?

Anyway, I need to start testing this out in the field to see if the signal will actually get through.