The Pocketbone project by the same guy is also worth checking out. It's a small Beaglebone based on the Octavo 3358, which is a huge SoC that integrates all of the hard parts of a minimal Beaglebone-like system. That project (also in KiCAD) can serve as a jumping off point for any embedded project where you need a tiny Linux system with hard real time performance and easy integration with high speed peripherals (thanks to the weird and wonderful PRUs in the AM3358 Sitara processor). These sorts of projects are really valuable in the embedded world because they demo a minimal implementation and in doing so reduce the magic that bogs down newcomers who just want to do something cool.
I'd love to see something 3358-based with the Ethernet connected to a male RJ-45 - Just plug it into the switch and have an instant cluster.
Extra credit if you can have it with power over Ethernet. Double extra credit if the Ethernet, power and one USB are available on an edge connector so a backplane with a simple switch like the Microchip KSZ8997 could easily be built.
The octavo chip is a part of a much larger change going in the chip world: the rise of SIP integration, which greatly reduced The barriers and costs of integrating chip sized systems , thus opening the field for new innovations and actors.
One interesting result from this is the pic32mz-da chip , a mcu with Rich peripherals , 2D gpu and 32MB Of Dram, all for $7.5/10k , and I believe this is just the beginning.
So basically, a full computer, inc. possibility of Linux and high level languages and libraries, at a price of an mcu.
I understand what you're trying to say, but on the other hand I've completed around 10 boards in the last year using only KiCAD and have loved the experience. I don't think I left anything on the table which would have been there had I used commercial software. But good point, i'm only doing basic boards.
There really aren't any good ecad packages for low prices (at least not that I ever heard of it). All cheap packages are just glorified 2D drawing programs (e.g. eagle, target).
Even some of the expensive EDA packages can be rather user unfriendly. I've just started a new job where we use Cadence Orcad so I'm having to learn that. It's not exactly intuitive.
I found all I used (eagle, kicad, target, protel) not intuitive to use and not very user friendly (which is said about almost every CAD package), but within even that group there are massive differences in capability.
The Pocketbone project by the same guy is also worth checking out. It's a small Beaglebone based on the Octavo 3358, which is a huge SoC that integrates all of the hard parts of a minimal Beaglebone-like system. That project (also in KiCAD) can serve as a jumping off point for any embedded project where you need a tiny Linux system with hard real time performance and easy integration with high speed peripherals (thanks to the weird and wonderful PRUs in the AM3358 Sitara processor). These sorts of projects are really valuable in the embedded world because they demo a minimal implementation and in doing so reduce the magic that bogs down newcomers who just want to do something cool.
I'd love to see something 3358-based with the Ethernet connected to a male RJ-45 - Just plug it into the switch and have an instant cluster.
Extra credit if you can have it with power over Ethernet. Double extra credit if the Ethernet, power and one USB are available on an edge connector so a backplane with a simple switch like the Microchip KSZ8997 could easily be built.
The octavo chip is a part of a much larger change going in the chip world: the rise of SIP integration, which greatly reduced The barriers and costs of integrating chip sized systems , thus opening the field for new innovations and actors.
One interesting result from this is the pic32mz-da chip , a mcu with Rich peripherals , 2D gpu and 32MB Of Dram, all for $7.5/10k , and I believe this is just the beginning.
So basically, a full computer, inc. possibility of Linux and high level languages and libraries, at a price of an mcu.
Brilliant!
Can't wait for KiCAD to become more mainstream. Projects like this will definitely help.
Great work guys, fantastic.
KiCAD is still missing basic features like via stiching. It's coming, but I can't wait for it to arrive.
Also, where is he getting the FE310?
Kicad is only viable if your time is worth less than money - basically only if you're doing the simplest kinds of boards.
If Kicad were as good as GCC, we'd have a revolution.
I understand what you're trying to say, but on the other hand I've completed around 10 boards in the last year using only KiCAD and have loved the experience. I don't think I left anything on the table which would have been there had I used commercial software. But good point, i'm only doing basic boards.
GCC didn't used to be as good as GCC is now ;)
So what do you recommend instead? I think KiCad is still not a bad choice for hobbyist use.
There really aren't any good ecad packages for low prices (at least not that I ever heard of it). All cheap packages are just glorified 2D drawing programs (e.g. eagle, target).
Even some of the expensive EDA packages can be rather user unfriendly. I've just started a new job where we use Cadence Orcad so I'm having to learn that. It's not exactly intuitive.
I found all I used (eagle, kicad, target, protel) not intuitive to use and not very user friendly (which is said about almost every CAD package), but within even that group there are massive differences in capability.