RAK831 error "Concentrator start failed: lgw_start"


My problem is regarding RAK831 migration TTN gateway from v2 to v3 and from packet forwarder to basic station. I’m using RPi 3B+ and RAK 831, all in Balena. All RAK831 worked fine with packet forwarder in TTN v2.
Then, I’ve migrated RPi&RAK831 to basic station in Balena as it is described here: https://www.balena.io/blog/deploy-a-basics-station-lora-gateway-with-ttn-and-balena/
It seems I was lucky and first GW working without any problem (!). Another three RPi&RAK831 couldn’t connect despite the equal settings. The error is the same:
[RAL:ERRO] Concentrator start failed: lgw_start

RAK 2245 instead RAK 831 works fine. Also tested with this settings in Balena, but doesn’t change anything:
SPI_SPEED 2000000
RESIN_HOST_CONFIG_core_freq 250

Despite problems with three RAK 831, fourth one in same environment works fine. Why three concentrators trying to connect all the time, unsuccessfully?

TTN console shows:
17:14:11 Disconnect gateway
17:14:10 Receive gateway status Versions firmware"1.0.0"package"1.0.0"platform"rpi - Firmware 1.0.0 - Protocol 2"station"2.0.5(rpi/std)"
17:14:10 Connect gateway

Balena’s log shows:
main 2021-07-20 15:18:51.856 [RAL:VERB] SX1301 ifchain 9: enable=1 rf_chain=1 freq=300000 bandwidth=3 datarate=50000 sync_word=0/0
main 2021-07-20 15:18:51.856 [RAL:VERB] SX130x LBT not enabled
main 2021-07-20 15:18:51.856 [RAL:VERB] Station device: /dev/spidev0.0 (PPS capture disabled)
main 2021-07-20 15:18:51.857 [RAL:ERRO] Concentrator start failed: lgw_start
main 2021-07-20 15:18:51.857 [RAL:ERRO] ral_config failed with status 0x08
main 2021-07-20 15:18:51.857 [any:ERRO] Closing connection to muxs - error in s2e_onMsg
main 2021-07-20 15:18:51.857 [AIO:DEBU] [3] ws_close reason=1000
main 2021-07-20 15:18:51.857 [AIO:DEBU] Echoing close - reason=1000
main 2021-07-20 15:18:51.904 [AIO:DEBU] [3|WS] Server sent close: reason=1000
main 2021-07-20 15:18:51.904 [AIO:DEBU] [3] WS connection shutdown…
main 2021-07-20 15:18:51.905 [TCE:VERB] Connection to MUXS closed in state -1
main 2021-07-20 15:18:51.905 [TCE:INFO] INFOS reconnect backoff 10s (retry 1)

Does someone have any idea where to start “witch hunt”?
Thank You for any hint!

Hello @Roberto69,

Welcome to the forum! This seem to be a a docker-compose permission “issue” to the basicstaion container like in this topic in balena forum: Problems with basicstation - #143 by mpous - Product support - balenaForums. You’ll need to create a variable called TC_KEY.

Can you follow the steps from the balena Basics Station repo here: GitHub - mpous/basicstation: LoRa Basics™ Station - The LoRaWAN Gateway Software to create that tc-key and check if the issue persists?

Hello @Martin
thank you for your answer. I’ve double checked and re-create variable TC_KEY, but it is the same.
Here are my variables:

I’ve also connect the same RAK831 to Raspberry Pi 0 and as GW it connects to v2 without any problem. I’ve tested this just to avoid suspicion concentrator is failed

Hello @Martin Martin it seems I have problem with RPi hat. What is interesting is, with packet forwarder in TTN v2 I haven’t any problem. Then I changed packet forwarder with basic station GitHub - mpous/basicstation: LoRa Basics™ Station - The LoRaWAN Gateway Software and migrated from v2 to v3, at the same time. One RAK831&Rpi 3B+ still works with occasionally “undervoltage” warning, but the other two not.
Is it possible, combination RAK831&Rpi3B+&TTNv3 drains more current? Sounds stupid but I have no other explanation … The same RAK831 on RPi0 starts, but on RPi3B+ doesn’t.

For future reference: @Roberto69 was able to solve the problem by setting the correct PINS to the RAK831:
reset pins

1 Like

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.