Sx126x support has any hope?

[addaedans]

Hi! Is it possible that there will be support for the SX126x(especially 1262) chip? There is example for ping-pong comm between these chips but it would be quite difficult to implement lorawan into it for me to say the least...The example codes are quite hidden but i can share the links if someone would take a look at it.

[terrillmoore]

I'm happy to look at the reference code, but I don't have test hardware, or know of a suitable source yet....

[terrillmoore]

@cstratton found the datasheet for me, thanks. It's not a big project to add support. Steps are:

1. make an abstract API at the radio.c boundary (the radio() function) -- this is just like the multi-region refactoring, use a macro to point to the real driver.
2. edit the current sx1272/6 driver to use a new name for the entry point
3. define HAL ways to get the radio configuration "as wired" pushed into the radio driver.
4. write the driver.

I don't think any changes are needed in lmic*.c for this. It's all radio.c -> radio_sx127x.c and radio_sx126x.c. I imagine some HAL changes are needed; possibly the whole pinmapping structure has to change because it's 127x-ish.

Next step is to figure out how to find the time (and the suitable reference platform)...

[addaedans]

Hi! Sorry for the late reply! Here is the link:
http://www.dorji.com/products-detail.php?ProId=65
At the resources tab there is the reference codes link which has all the files for the ping-pong comm.
Hope you can get it working :)

Edit: oh i see you found it...well that seems others are interested as well ;)

[addaedans]

Hi! Is there any progress in the support of the 126x chipset?

[terrillmoore]

Nope, I don't have hardware, and I don't have budget... So unless we can get someone to offer MCCI support to do this, or someone else wants to do the work, it won't happen for a while. Sorry.

Best regards,
--Terry

[addaedans]

Hm, i think maybe we can sort this out. If i buy you two 1262 chip and send to your address it could work? Terry Moore <notifications@github.com> ezt írta (időpont: 2019. máj. 8., Sze 20:09):

…

Nope, I don't have hardware, and I don't have budget... So unless we can get someone to offer MCCI support to do this, or someone else wants to do the work, it won't happen for a while. Sorry. Best regards, --Terry — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#301 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGKX7SMCQRUMPQAOR44BO23PUMJOLANCNFSM4HEDZWRQ> .

[terrillmoore]

Chips aren't sufficient, alas, we need working systems (CPU + chip + RF, etc.). There's also the issue of funding the engineering work and ongoing support. The cost is likely to be out of reach for an individual, but reasonable for a company. It could be crowd-sourced, I suppose, but I don't have the time to administer that kind of thing; and we spent my pro-bono budget for the year already on pi64.win.

Drop me an email at tmm@mcci.com and we can discuss further if you're interested.

[slavendam]

Hi,
any news about this?
As I can see, SX1262 became more available and suppliers sending notifications that SX1272 will become outdated soon (not sure what that "soon" really means).

[terrillmoore]

@slavendam: Technically very possible. No funding. No notifications from MCCI's suppliers about the SX1276, which is what we use.

As said above:

There's also the issue of funding the engineering work and ongoing support. The cost is likely to be out of reach for an individual, but reasonable for a company. It could be crowd-sourced, I suppose, but I don't have the time to administer that kind of thing.

I can easily use the Heltec modules for reference. But without a business case to do this, my time is not available. I had no response to the quoted comment previously, so I assume that there is no commercial interest in this. MCCI can't fund it, and I can't fund this myself; the pandemic era has been challenging. Sorry to disappoint. If someone wants to volunteer to run a fundraising drive (via The Things Network NY, which is a non-profit), then it could be crowd sourced. But I don't have the time to run a crowdsourcing program.