update all boards to use new radio standard init

[446564]

Discovered while working on fixed bluetooth pairing code configuration.

At least on the nano g2 ultra, when this clause is included the device never completes boot, just hangs after display init with Please wait... on the screen. I traced commits until I found this causes the issue and confirmed on latest dev it works without these lines and fails to boot with them.

https://github.com/ripplebiz/MeshCore/blob/7f0f3b7753fbec94dfe7a36369a6ab7a167dff77/src/helpers/CustomSX1262.h#L47-L51

Device Variants to be Updated

• heltec_ct62
• heltec_tracker Refactor radiolib for Heltec Wireless Tracker #462
• lilygo_t3s3 refactor: lilygo_t3s3 with CustomSX1262::std_init() #464
• lilygo_tbeam_supreme_SX1262 T-Beam refactor for radio.std_init #427
• lilygo_tbeam_SX1262 T-Beam refactor for radio.std_init #427
• lilygo_tlora_c6
• rak4631 refactor: RAK4631 with CustomSX1262::std_init() #429
• waveshare_rp2040_lora
• nano g2 ultra update nano g2 to use radio.std_init #428

[oltaco]

That's strange, it shouldn't be even getting into that if statement there unless the radio init failed in the first place. Without this check for failed radio init and subsequent attempt with txco set to 0.0f we can't support both HT-RA62 and RA-01sh modules in the same promicro target.

Can you apply the radio.std_init() refactor for station g2 and see if that fixes it?

[446564]

I'll check that out shortly, it's certainly odd because without this it boots and then works as expected.

[446564]

Problem solved, so I guess we should update all variants before closing.

[446564]

do we think this also affects SX127x? sounds like maybe heltec are experiencing the same issue.

[oltaco]

do we think this also affects SX127x? sounds like maybe heltec are experiencing the same issue.

Don't think so, there haven't been changes to those wrappers as yet.

[446564]

do we think this also affects SX127x? sounds like maybe heltec are experiencing the same issue.

Don't think so, there haven't been changes to those wrappers as yet.

I didn't think so but saw users reporting similar issues with the heltec but a reboot cleared it which was not the case with this issue. So something else, happens only just after a flash.

[jbrazio]

Do not break other boards..
That code seems to be a fallback in case the board has no TCXO or the wrong value was selected.

https://github.com/ripplebiz/MeshCore/blob/478a57a6bdab2e22a7e5b8cb7bf445f706294fe5/variants/waveshare_rp2040_lora/target.cpp#L37-L38

You can see the radio is init using tcxo which is set here:

https://github.com/ripplebiz/MeshCore/blob/478a57a6bdab2e22a7e5b8cb7bf445f706294fe5/variants/waveshare_rp2040_lora/target.cpp#L22-L26

and MACRO defined here:

https://github.com/ripplebiz/MeshCore/blob/478a57a6bdab2e22a7e5b8cb7bf445f706294fe5/src/helpers/rp2040/WaveshareBoard.h#L19

The board which you are having the issue might have the wrong MACRO value for SX126X_DIO3_TCXO_VOLTAGE ou using the default value of 1.6f. I don't know if all variants are actually setting the same default value as this is depended on the helpers/**/*.h files.

[446564]

Do not break other boards.. That code seems to be a fallback in case the board has no TCXO or the wrong value was selected.

MeshCore/variants/waveshare_rp2040_lora/target.cpp

Lines 37 to 38 in 478a57a
int status = radio.begin(LORA_FREQ, LORA_BW, LORA_SF, LORA_CR, RADIOLIB_SX126X_SYNC_WORD_PRIVATE,
LORA_TX_POWER, 8, tcxo);

You can see the radio is init using tcxo which is set here:

MeshCore/variants/waveshare_rp2040_lora/target.cpp

Lines 22 to 26 in 478a57a
#ifdef SX126X_DIO3_TCXO_VOLTAGE
float tcxo = SX126X_DIO3_TCXO_VOLTAGE;
#else
float tcxo = 1.6f;
#endif

and MACRO defined here:

MeshCore/src/helpers/rp2040/WaveshareBoard.h

Line 19 in 478a57a
#define SX126X_DIO3_TCXO_VOLTAGE 0

The board which you are having the issue might have the wrong MACRO value for SX126X_DIO3_TCXO_VOLTAGE ou using the default value of 1.6f. I don't know if all variants are actually setting the same default value as this is depended on the helpers/**/*.h files.

Nothing is getting broken.

That fallback code is part of the new standard initialization process and is not being removed with this issue.

Each board can define it's own tcxo voltage which is respected as seen here: https://github.com/ripplebiz/MeshCore/blob/478a57a6bdab2e22a7e5b8cb7bf445f706294fe5/src/helpers/CustomSX1262.h#L18-L20

[seagull9000]

Heltec Tracker PR #462

[oltaco]

PR for Lilygo T3-S3 @ #464