After four months of continual silence from life getting ahead of me, I have had a chance to sit down and truly figure out what the platform will be based on (who got knocked out by me being in the quarter finals even though I resigned?). I had my original mind stuck inbetween a particular ARM SoC and Intel Atom SoC. And, with much deliberation and one trip to see how easy it is to purchase each of these processors it was an easy decision. Intel seems to be the winner right now by a land slide. This is because I can purchase most intel processors, scan the bottom and map out the bga on my ridiculous scanner, and I don't have to go through a paper wall to get the processors. Not one ARM Cortex-A model was quickly available from any of the manufacturers. Well, at least I can put all my ARM assembly into the Raspberry Pi. ***I should have been focusing on *nix like Kernels***
Things to do now is send my power system design to OSH to get made, and have it sent back with a screen so I can do the soldering here (I DON'T HAVE TO USE MY TRANSPARENT LASER PRINTER SHEETS FOR SILKSCREEN :D). I need a test subject for the new reflow oven, and three of those boards would be great.
Write 2 versions of the drivers to control the main power control on the board. One practical prototype for the Pi, and one for the x86_64 design (If I hadn't looked at ARM I would have went for x86_64 sooner.... this is madness... we all go a little mad).
Next up I need to use my Pi as a test dummy to run the embedded USB 2.0->1.0 Hub and the power supply. I need to make sure that those things run correctly, and I might run video if I want to make a sample board that is 4 layers that separates power, signaling, video, and more.
Once I have all that working, and I have proper power system I can tackle the real beast. the 4+ layers of the final power prototype, and the 4+ layer mainboard prototype. I will be sacked studying every detail for a few months to a year. My lab* at that point will have to be rebuilt from the ground up. There will be minimal margins of error, and every error will easily cost me weeks.
Also, pictures, video (4K), "documentation", and giveaway prototypes when all is said and done (depending how much I hate the project when I get to my goal... More I hate it the more likely I will give away prototypes and not make more). If, I give away prototypes they get sent to these people first.
My friend who helped me get started by building my initial lab with good equipment.
Myself *why so low on the list me* -because you build bad things and should feel bad
Now, I can get down and refine what my image on what the specs have a good chance of being.
Power system will be based on a TI chip, and will be controlled by saved settings in the OS. The device defaults everything to on until a boot sequence tells it otherwise. This controls the distribution of power to all the individual segments on the device, it can't control the power of the main board, switches between charging and charged circuitry, and performs the first power on test then passes this onto the main system. That way you can upgrade and replace the main system without the power on sequece being completely in control of the main device. This also informs the main computer if a device is disconnected after all power has been removed from that device.
*Also, the charging port will be on the oppisite side of the board of the main control logic of the power/video/usb.
Each segment will have each device connect via a friction connection (think game boy, sega, and nintendo cartridge). This seemed like something that would last a considerable amount of time, be simple to produce, and easy to use. All other solutions required too many tools. Power will be turned off by definition of the user. Not only will there be usb connections...
Read more »