[Arm-netbook] Release of SoM board KiCAD files with 2x100 pin ARM as a stepping stone

[luke.leighton]

On Mon, Dec 10, 2012 at 3:54 PM, jm <joem at martindale-electric.co.uk> wrote:

> 1. Luke, I have pinout of BGA for A10
> http://www.gplsquared.com/SoM2/SoM2.html
> but it is not accurate. Many of the NC pins have functions when I check
> against Cubies design, you design and Wits design. I'm not lost, but we
> need a full BGA chip pinout diagram to avoid getting lost. I can reverse
> engineer from all the different drawings - but if someone has it
> already, share it please!!

 that allwinner.lib i created is the only one i know of, and i made
it.  yes the datasheet has missing stuff, yes i too added things as i
found them.

 btw WATCH OUT.  kicad (including latest version) does NOT allow
successful editing of the allwinner.lib file - 441 pins seems to be
too much for it.

 luckily i've been able to do editing by hand - text files.  good eh?

> 2. The wiggly lines issue. Personally I believe utter cretins are
> stitching up everybody with this one. I have university degree in
> electronics and I just don't believe any of it because it doesn't agree
> with any of my training of high speed electronics.

 *cackle*.  yes - the app notes i've read all do say "for god's sake
don't make a square track" because the signal speeds are approaching
R.F. and they tend to keep going in a straight line.  wiggly
combinations of 45 degree bends however are a compromise that seems to
work, empirically.

> So what I propose is that I design the layout as it should be done.
> I have budget for several iterations to get it right.

 awesome.

> 3. The ohms thing is relevant but lots of ways to mitigate it - KiCAD
> the latest edition has calculator for that.

 oh good.  that's really good to hear.

l.