Is SiK firmware maintained?

Is there someone who maintains the current SiK firmware upstream? The original SiK firmware repository seems to be unmaintained for few years.

We plan to expand the SiK firmware by two antenna MIMO and some more features. We already ported the firmware to the newer Si1060 chip, which is an improvement by itself, because the newer chip has better sensitivity and new improved coding methods. But the pull-request is without any reaction for almost half of the year.

  • Have I missed something and SiK firmware was replaced by something better?
  • Are the SiK firmware users satisfied with the current state of the development and no new features are needed?

Is there some possibility of how can I help to resolve that? My current personal opinion is the SiK firmware is still relevant for UAV telemetry radio link.

1 Like

I’ll give an ironclad guarantee that the maintainer of that repo won’t be monitoring this forum, but he’s very accessible in the ArduPilot discord.
I’d suggest reaching out and pinging him, as it’s likely the pr just got forgotten. Perhaps clean up and squash your commits too :slight_smile:

Are these easily available to purchase anywhere online?

Not yet. Now, it is in the state of prototypes. But we plan to sell it publicly. Now we need to adapt it for use on UAVs (add JST-GH connector, do it more compact and include input/output for two antennas).

There are some technical difficulties that complicate the connection to the autopilot or to the USB. As different levels or some logic for disconnecting of USB-UART convertor. Commonly available modems not have solved that. And 5V logic is used for 3v3 autopilot logic (according to DS-090) but 5V tolerant (in case of CUAV autopilot)

I don’t want to be a party pooper, but remember that this is a RF device - you need to obtain all the legal certificates (CE, FCC and other depending on country) to be able to legally sell it.

Yes, we are aware of that. The main question of this topic is - Why is the SiK firmware stuck in several years old code (with many unsolved pull-request) and also why the development of new hardware modems based on new components seems to be frozen.