Hey everyone! Time for another update, sorry it's been a few days since I posted anything substantial.
Progress has been slow for the past week thanks to unplanned long work hours at my day job and some family stuff, but I finally had the time tonight to continue with the prototype and managed to get it working and injecting a payload to my Switch! Took a bit of time and troubleshooting to figure out a way to flash the bootloader to a fresh from the factory ATSAMD21, but in the end a Raspberry Pi Zero with OpenOCD worked best after some learning on my part. As it turns out, the ST-LINK V2 I bought can't be used to program the ATSAMD21 unless you flash it to a third-party CMSIS-DAP firmware, which ironically requires a second ST-LINK-V2 to flash the first, so the Pi won out. The Pi will also make it easier to streamline flashing large amounts of these quickly since I can turn it into a sort of push-button-flash-firmware device with some pogo-pins, a push button and a 3D printed jig, with no computer needed.
Once that problem was solved I ran into another one, which is that the dongle stays powered all the time, meaning the MOSFET isn't doing its job. I'm not entirely sure why this is the case, since I'm pulling the gate down properly with a 10K resistor and a diode prevents any voltage from anything but D+ or USB VCC. I checked the gate to make sure it was getting triggered only when it was supposed to and it was, so I have no idea what the issue is. In any case, I subbed in a transistor with some careful soldering and it worked great even if it looked pretty ugly. Unfortunately this means a V4 board is necessary, which I'll be ordering tomorrow, and sadly adds another week until testers go out.
On the plus side, I did some tests with moving the LDO output past the current limiting resistor and it works great, so I'm incorporating that into the V4 design for vastly quicker USB charging times similar to the SX Pro dongle. Everything else on the board seems to work well including the RGB LED, so while the delay due to the MOSFET issue is unfortunate, I'm still happy with the current progress. As for the second button for payload switching, I'm going to leave it in, since the new multi-payload version of Hekate reportedly doesn't work with the SX OS payload yet. Plus, I can always re-purpose the button for something else later.
Also, I spent some time tonight practicing my hand soldering on some V3 boards, and while the USB Type C port is a bit tricky, everything else is pretty easy to solder by hand, including the ATSAMD21. I originally was soldering everything but the Type C port and the supercap with my hot air station, but it's actually easier to just solder it all with a fine ball tip on my TS-100. Drag soldering ended up working really well for the ATSAMD21 and takes very little time. I'm going to add 0.5mm of length to the Type C pads to try and make it less tricky to solder. I don't plan to hand solder more than the first couple of batches of these if I can help it, whether I get help from a PCBA service or get myself a part picker, but it's nice to know I should have no problem getting a good 10 of these done a night.
Anyways, just wanted to let everyone know progress is still happening and none of my plans for the project have changed, even if the timeline has been pushed back a bit. As always, I welcome any feedback you guys have. Thanks again for all the support!