Ok, thanks! Seems as if you are correct! I am now using the following 3 gpio's: 2, 3, 18. Seems to be the combination that works for me. Would be nice if we could get some sort of definitive doc that might let us know whats up. Would also be nice if the docs were consistent.
I tried a number of solutions. Here’s what worked and what didn’t. “Worked” here means that the Omega2 booted correctly and then successfully drove the DotStar LED strip via the SPI interface using Python code linked above.
I found that I had to connect the output enable pins of the 74AHCT2G125 to ground. That surprised me, as I expected that they would have to driven by the Omega2 SPI CS1 pin, allowing the Omega2 clock and data pins to float while the Omega2 was booting. But with the OE lines soldered low, it worked, booting and then driving the LEDs via SPI.
I soldered the pins of a mini-expansion board as described in the original post above, wiring the +5V from the DotStar power supply, via the connector cable to the 5V pin of the USB A socket on the mini expansion board. It's large enough to be probed/soldered to. [Note: With this setup, power is now supplied by an external supply. Do not connect to a laptop using either USB port.]
I sent this question to Onion tech support, but received only a note that it would be forwarded to technical staff.
it looks like the Omega2 headers are labeled "A" and "B" and numbered left to right. Diagram here
The open source mt76 driver in LEDE master has received some bugfixes today that apparently make it work reliably now! At least my current testing indicates no more need to mess with kernel version dependent proprietary binary blob drivers - mt76 for the first time just works stable! #knockonwood
The same as @Andrea-Cesari: mine gets pretty warm within first minutes of launching and keeps the same temperature independently of the CPU activity - but I can easily hold my finger on it, it does not get jalapeno-hot.
@Ismas the architecture (target) in menuconfig is called "Ralink RT288x/RT3xxx", under which you find a subtarget "MT7688 based boards" and finally the Omega2 profiles.
Before starting to dig through Linux sources for these SoCs I hadn't realized how difficult it is to relate chip names to their architectures and drivers. There's little logic in it, but a lot of history.
How would one be able to guess the MediaTek 7688 goes under the "Ralink" architecture? Of course, once you know that Ralink was once a company that made MIPS based SoCs, and was aquired by MediaTek in 2011, it makes sense.
Still, support for the MT7688 can go by any names with "ralink", "rt", "mt76" "7621" and probably quite a few others. Comments in drivers often refer to subsystems that probably once were separate chips and thus had a completely different number. And all this being in the mostly proprietary SoC world (few proper datasheets) makes it hard to figure out what belongs where. So it took a while for a beginner like me to even start connecting dots...