I have now completed the assembly phase of this project. While almost everything ended up working as planned, the process wasn't without its problems. Soldering was more or less routine, that was fine. The semi-permanent, sandwiched-PCB design of this device proved to cause problems with debugging and testing though.

The first issue was with the FM transmitter, as it was not transmitting at all. Everything else was working fine, I2C and everything. After testing with an external antenna circuit, I was able to narrow the problem down to a defective SMD capacitor in the on-board antenna circuit. I guess that's the risk you take when using very cheap parts from aliexpress. After a new capacitor and some software reconfiguration for the lower frequency crystal that was used, the FM transmitter worked perfectly.

The next problem was with the USB hub circuit. Originally, I thought it was just the schematic that I used for the USB hub that was bad, since I did not actually prototyped anything for USB before ordering the PCBs. I did not yet have the necessary USB IC, and waiting for it any longer would reduce the time I would have to reorder the PCBs if any large problems arose. However, after testing the same circuit on a breadboard, everything appeared to work fine. I then started disconnecting parts from the breadboarded circuit like capacitors and resistors, until it stopped working. In my case, a missing 2.7K resistor on the REXT pin of the FE1.1s is enough to stop everything in its tracks. I checked the same resistor's connections on the problem device, and it turned out to have bad solder connections. Fortunately, the resistor was very close to the edge of the device, so resoldering it would not be a problem. Otherwise, I would have had to desolder the Pi to get to it, start over and desolder the Pi anyways, or just give up on the USB hub. After that simple fix, everything with USB was good. Stupid me managed to break off a USB port during testing, but it did not damage anything else and there are still two good ports left.

The last problem was me overestimating the brightness of the power LED on the front. Those tiny green LEDs are blindingly bright, and I didn't really want to blind myself while driving at night. The solution was a simple resistor replacement, which then made things a lot more tolerable.

Overall, I am very satisfied with the final result. The overall design can still use some improvement though. For v2, I plan on making some of the following changes:

• Add a light sensor and a good PWM output for LED control
• Add audio input and output jacks like previously planned
• Try to find a better form factor display
• Remove some unnecessary things like the serial pins, etc
• General hardware arrangement

Programming is coming next.