Thanks! I just don’t know how to troubleshoot it further. I have to say I was testing with TTN V3 - I’ve now moved back to V2 as I didn’t realise how people weren’t yet migrating to TTN V3. I’ll test and report back. I’ve not got much to go on right now as I’ve tried following the code to find out where it’s going wrong, but it’s rather tricky to get to grips with!
OTAA not working with LMIC often indicates that LoRa (D)I01 is not connected or not correctly mapped. DIO1 is required for downlinks (which includes OTAA).
(D)IO2 is not needed for LoRa (only for FSK).
Uplinks only need (D)IO0 which may explain why ABP works but OTAA doesn’t.
On the T-Beams both DIO1 and DIO2 are already wired on the PCB so you may have a pin mapping configuration issue.
Another option for not being able to join is when there is no gateway coverage in the area.
Using ABP the device will just do it’s thing even without any gateway (but the uplinks will never arrive). With OTAA without working gateway connection your device will never get a join accept. But this is not relevant in your case because you see messages arriving in the console when using ABP. R
FYI: On the TTN forum you can find much ESP32 and T-Beam related information and tips.
The following can probably help to get your T-Beam working with OTAA: LMIC-node
I agree with @lnlp that it seems your problem is likely to be specific to your board.
One of the (many) things on my list is to come up with a step-by-step bring-up procedure (with suitable software support from the LMIC); and a proper logging system (so people get non-disruptive logs, and tben post them). Not likely to happen soon unless funding/resources magically appear. (Wish I could figure out a crypto-currency/NFT angle to all this!)
BTW, I’ve been using TTN V3; it’s working well for me (apart from the fact that the 5 second delay broke another library… but the LMIC works fine). I don’t think that’s your problem.