While parts are on order, might as well spend some time on the mainboard. Over the past week and a half, I've been tweaking trace placement to increase contiguous ground planes while removing vias, test fitting components and adjusting footprints, increasing the thermal relief sections of copper for the supply rails, and also fixing clearances.

Oh, and realizing that I forgot to put a pull-up on the MCLR pin, decouple it, and put a current limiter on it; important, huh? Speaking of semi-important things, the pin header footprints were revised for trace space and they were all shuffled so a jumper cannot be incidentally placed in a way that shorts a pin and so silkscreens can be read more easily.

I've removed more than a dozen vias for creating or jumping ground planes together but I've added half of them back to prevent any plane from behaving like an antenna. I noticed as well that clearances were 6 mil and though this meets many labs' capabilities, it doesn't make it cheap to manufacture. I've since adjusted everything sans 2 vias to 7 mil clearances, though I still have 6 mil traces in a number of areas. The 2 that fail are literally less than 1 mil off so I'm not fretting over it. I can increase trance widths to 7 mil, but will take more time than I'm willing to spend on it since it'd require reworking the MCU-SRAM section of the board.

The thermal relief areas for the 14V LDO, 5V LDO, and 3.3V SMPS were bolstered. When going over the LED power requirements, even with PWM, for the display boards, I saw that I was going to be cutting things close. The other issue I do not know if it'll be an issue is the EMI created by drawing 300-450 mA @ 14 V for about 10-30 us. All power traces were upped in size a bit too. I want a low of a temperature rise as is reasonable for the board.

Anyhow, I have added 3 fiducials on the board, both front and back, and they are on the top of the board and the bottom right. The top pair share the same Y-axis coordinate and the right pair share the same X-axis coordinate. Don't know if I'll ever pass it through a pick and place, but it didn't hurt adding them in. I happen to push the mounting holes close to the corners. Even if these are cutting off the corners, which they're not, they'll provide notches for mounting. Corners were rounded for "workability" as when this gets stuffed in a dashboard, wires are everywhere they'll help getting hung up.

Anyhow, I'm dubbing the current revision as prototype 2.3. I'll let it sit around for a few days and see if there is anything else that jumps out at me. Though the next week or so is dedicated to soldering and getting firmware loaded on the MCU, due to the oversight of the MCLR pin, I'm not sure how it'll behave. Thankfully, I can still place a PU resistor on the board, but it'll be a ways away from the MCU.

My last possible show stopper is the mDP port. It's a reversed, aka upside down, port so the odd and even pins are swapped. As this is a fundamental change to component that I did not realize, I need to verify that I'm not sending a signal down a ground/shield path and that I'm on the correct copper pairs so I know where to expect a signal to show. Once I have this sorted, I'll order up the latest prototype from a lab.

Proto 2.3 Front:

Proto 2.3 Back:

h, forgot to note, I've almost settled on an appropriate license for the board. Through the deliberation, it's apparent that if someone every will want to rip off the board, the license is irrelevant as that individual or group doesn't care about it. The license will only limit myself and those who want to use it while maintaining its integrity.