You're right, I wrote a bit of a mess. I am not currently working in a company and I want to improve myself in hardware Design field. What I really want to learn is what path should I follow while developing myself? I follow the trainings, but I want to do projects as if I was working in a company, but I am inexperienced. At this point I don't know how to proceed to the next step.
Well doing a whole board as a DIY project is very time consuming. The big problem with starting out like this is that you do not benefit from working with seasoned engineers. However the internet does make up for that in part. You can get serious help though you will need to keep the scope of the project under control, to that end I'd suggest a small single board controller with the board no bigger than Arduino or maybe Raspberry PI. Put some high speed ports on it and maybe even some ram, I'm thinking SoC so that the ports would be available in the first place. The proceed to get everything working. By the way there is a huge cost in software here, almost every high speed port these days requires serious software support thus you seldom see one man success stories. If there are success stories it is usually because they purposefully narrow scope. You may need drivers or whole operating systems ported just to get hardware to work.
I'm always scratching my head when it comes to many of these single board controllers as I'm always asking why didn't they do "this". Cost is often sited which may be true but a DIY project will not suffer from this. So why not consider such a board with a really fast interface to an SSD of some sort. That all by itself should offer some high speed challenges. Plus Linux would really benefit on such boards. If you go with a SoC, with built in RAM, this should easily fit on a small board.
By the way yes I know those SoC are expensive but they also come with the high speed ports that need to be proper;y routed to port connectors. Documentation may be hard to come by too. However what yo will end up with is a variety of high speed ports from USB to video to whatever. Since this seems to be what your are interested in an SoC that supports many of these sorts of ports seems to be the right avenue. Plus if you want to have something that shows off skills well modern hardware is the place to do it.
As for how do you spec something like this, build it so that it has what you want in small single board computer. If you want two USB ports and a modern video port than design them in. If you need a dog slow serial port design them in. If you want a fast SSD then design it in. The trick is to come up with the specs before hand and then try to pack everything in. I'm thinking a board here capable of running Linux and that simply because there is an "easy" path to hardware support to actually get the board running.
On the flip side there are lesser micro controllers that may have enough software support from the manufactures to allow you to do a simpler board. The idea here is that they would have software to support high speed USB and maybe a port to an SD card. Again the goal is to minimize the need to write a lot of low level code to get your high speed ports to work correctly. The problem here is that there might not be much in the way of high speed electronics to support / design for. Just avoid another Arduino compatible board, too many of them exist. From the design standpoint though the effort is the same, you need to define what you want on the board and go about supporting that functionality. Every bodies needs are different, I might say I want two old fashion RS232 ports and the next guy would laugh at that and say he wants CAN bus support, and the guy next to him would want 4 USB ports. In the end the Microcontroller might define what goes on the board. Sometimes the specification is the hard part.
In a nut shell we can't tell you what to design. High speed can mean different things to different people, you could have ideas far beyond what I expressed here. In any event I'd love to see you bring a nicely designed ARM board to market. Why go work for somebody else?
Home
Help
Search
Login
Register
