[Arm-netbook] [eoma68] mini desktop pc

[joem]

On Tue, 2013-12-03 at 11:35 +0000, luke.leighton wrote:
> On Tue, Dec 3, 2013 at 10:00 AM, joem <joem at martindale-electric.co.uk> wrote:
> 
> > 1. You need the internal 3.3V line brought out from EOMA68 to one of the
> > pins for general circuit design reasons.
> 
>  not going to happen.  sorry. designs need to take that into account.
> apart from anything there will be boards in the future that don't
> *have* a 3.3V line (for whatever reason).

Not a request.
The 3.3V line is the pull up line for GPIO lines.
Something to get right in next revision.

> > 2. Not sure about the USD75. Is it going to get cheaper any time soon?
> 
>  in quantity 10k and above, probably yes.  because they will be
> designed differently.  and be a different product entirely.  so... no.
> 
>  plus remember that $75 is for *two* boards, and tax, and shipping,
> and testing, and packaging, and profit (a small one) and everything
> else.

Its $75 for CPU board and meb + shipping + taxes.

> 
> > 3. At the moment there is no way to cooperate with everyone using
> >    closed sources PCB packages and the list of hardware mistakes
> >    are piling up as cost somewhere. This is driving a wedge through
> >    all things good and possible. I only want to work in open sourced
> >    KiCAD. The limitations are not significant
> 
>  they are a complete show-stopper for someone with my limited
> available time, lack of knowledge and insufficient expertise.  simple
> as that.

KiCAD web site has added a lot more video tutorials recently.
Watch it 10 times and follow blindly until it ground in.
I can also help make the 3D parts and answer a lot of questions
until its all up to speed for anyone wanting to do all this.
Searching the kicad mailing list also answers many a question.

> >    to getting the real work done and avoiding undesirable future costs.
> 
>  then i need people to take over designing these I/O Board PCBs.

The main 'problem' with kicad is that it is too fast once you know how
to work it.
Unlike the commercial packages, all the data is stored in text files.
Simple diff after a modification will show which parameters get
modified when a small change is made.
A lot of the difficult work can be done with text editors,
spreadsheets, and copy/paste which speeds up the work enormously.
You end up with dozens of boards ready to go and not enough time
to test.

>  as long as i am continuing to work without any visible day-to-day
> assistance i will continue to use the software that i have the
> capability to use and have confidence in.

Board design has to be done in small incremental steps.
Just like unix philosophy, build it small and efficient and make sure it
works 100% and fully debugged before moving on to the bigger things.
So the real work is to produce an axp power supply module, an ethernet
module, a DDR + CPU module, audio module, flash module, lcd module, hdmi
module, vga module, etc and making sure each bit works 100% and then
adding them together to make final product. The modules cost less and
easier to fund. Once modules are working, you never look back and
race to build bigger and bigger projects at a faster rate than
anyone else can match.

Monolithic designs are great if you have resources to make
additional revisions and to pound each variant with
enough resources until it works.