July 20, 2025
Welcome to part 3. I had the boards manufactured by JLCPCB at the start of this month and they arrived yesterday. They look great!
Then I went to plug one into my computer and it didn't work. No enumeration on USB in boot mode and, as I would find out later, not even any response over SWD.
After a little research I learned that the Hardware design with RP2350 document leaves out one important component: a pull-up resistor on the chip's RUN line. Without this the line either floats or is pulled to ground and the chip won't boot.
I added the resistor and it still didn't work. All the voltages are correct, including the 1.1V produced by the chip itself (I was worried about this because the regulator and its inductor are very sensitive to component placement). There simply was no indication that the chip even worked. At this point I put the whole thing down for a while.
I only had JLCPCB populate two boards, because populating all five took away my qualification for the really cheap shipping. This morning, I grabbed the other board, added the pullup resistor, and plugged it into my computer with the BOOT button held down---and it showed up! This means that not only are all the power circuits correct, the crystal oscillator circuit and layout is also functional. A little bit of code later, and:
It works! As of me writing this post, this is the only working Sharpie board in the whole world. I think that's pretty cool.
The program that's running is a super simple blink + write to USB serial bit of code. Neocities doesn't let me upload videos without paying, but if it did I would show you the blink.
I also placed a Digikey order with JST connectors for the debug pads, display connectors, a few accessories, and the displays. The displays are now listed as coming back in August, not October, so I had Digikey delay the whole shipment until they're in stock.
I see now that soldering the connectors by hand may not actually be possible, but I'm going to give it a shot anyway. I have three blank boards to practice on so I'm confident I can get it figured out.
That's all for now! I'm going to try to figure out what's wrong with the other board. It's probably either bad soldering (JLCPCB is not known for exemplar quality) or I actually managed to destroy the chip by powering it with RUN improperly configured (feels unlikely but anything is possible).