New Life in Openmoko Phones

[Nils Faerber]

joakim at verona.se schrieb:
> Nils Faerber <nils.faerber at kernelconcepts.de> writes:
>> Wolfgang Spraul schrieb:
>>> Today Openmoko released additional pieces of documentation about 
>>> Freerunner hardware: board outline, footprints and netlist.
>>> Same as all other releases before - under Creative Commons Share-Alike 
>>> license.
>>> Available at:
>>> http://downloads.openmoko.org/developer/schematics/GTA02/gta02_outline_footprints_netlist.tar.bz2
>> This is in general great!
>>
>> But sorry to be a little sceptical here - but hardware != software. What
>> I mean is that collaboratively developing software is pretty easy since
>> we have the internet to share and most of us have a PC to develop upon.
>>
>> But with hardware development the situation is a little bit different.
>> Even if the collaborate development effort succeeds, i.e. KiCAD is
>> sufficient and a hardware design becomes ready, it still needs to be
>> produced - and here troubles start, from buying the parts, making PCBs
>> etc. running up the whole stack to asembling the whole device and
>> testing it. This cannot be done as open source effort with volunteers.
>> Here real money is involved - a lot of real money. And this needs to be
>> done several times, for prototypes, small A-series, probably a B-series
>> and then final devices.
>> But you should know better than me about this process (at least by now).
>>
>> What are the plans or ideas to enable later on production?
>> Pleas eget me right, I would love to see such a project succeed and
>> maybe even contribute to it but I really cannot imagine any possibility
>> how such a hardware production should work in the end without a big
>> sponsor in the background.
> 
> It was many years since I did any serious electronics work, but from my
> uninformed viewpoint this seems to be workable because:
> - The case is not changed and can be reused
> - no parts are changed so existing inventory at OM can be used for
> prototyping

This would be one of the details I am interested in, i.e. would OpenMoko
Inc. help in making (read as "producing") this new design? With its part
stock, manufacturing capabilities, etc.?

If this can be answered with "yes" then this project is a *great*
opportunity.

> I dont know what making a PCB and populating it costs these days, but if
> it costs a couple of hundred euros per populated board, I would sponsor
> at least one out of my personal curiosity. I used to be good at
> electronics assembly, maybe I could even put it together myself if I
> find a SMD oven etc.

He ;)
Many of the parts in the GTA02 cannot be reasonably placed by hand.
There are almost a dozen (or more?) BGA chips which are extremely hard
to handle (you do not see if the balls match the pads). Then there are
almost microscopic parts like resistors and capacitors - which pitch?
0402 at least if not even 0201 or smaller. So populating the board is
almost impossible by hand without highly qualified tools (and no, a
tweezer and a stereo-microscope will not suffice).
But the problem you will encounter beforehand is printing the solder
paste. Stencil printing such high density with even and correct paste
distribution is not exactly easy even if you have proper stencil
printers. Adjusting them, having the right paste to print etc. is high
art of SMT manufacturing. And finally you need a really proper nitrogen
flooded full convection reflow oven for good quality soldering of such
delicate parts (different heat absorption of parts, proper heat
profiles, good energy distribution, etc.).

So what you really need is a modern manufacturing line, with auto-placer
for almost all parts. I do not know how many different parts there are
on the GTA02, probably 100, or even more? This means very high initial
effort for setting up the machine to pupulate a board. If you then run 1
or 10 or 100 does not make much difference for the machine cost anymore
(you just need more parts). The setup effort is the thing that makes
prototypes or small series such expensive. I just visited another
electronics maker here in Germany and they have a placement machine
which can set up to 85000 parts per hour. Compared to something like a
day for setting up the machine, the time placing the parts is almost
irrelevant.

The smaller the parts have got in the past the more difficult it has
become for hobbyists to catch up with technology. It will not take very
long until home-grown PCBs will be almost impossible to do because all
the interesting chips come as bare-die only (just the silicon, no case
or pins).

So what is needed is the real commitment by some professional hardware
manufacturer to put the new design on one of his lines and care for the
prototyping and small initial a-series. After the design has proven to
work a small first production run should be easier to setup since you
can then give proove that it will work and persuade potential customers
to pay up-front for the device - or at least a part up-front. That would
enable buying the parts and paying for setting up the production. I
think the Open Pandora people did it quite similarly, i.e. they sold
devices and had them made after sales. If your customers trust you
enough this can work.

So in the end hardware making is more a matter of money than motivation
or man power, pitily...

Cheers
  nils faerber

-- 
kernel concepts GbR        Tel: +49-271-771091-12
Sieghuetter Hauptweg 48    Fax: +49-271-771091-19
D-57072 Siegen             Mob: +49-176-21024535
http://www.kernelconcepts.de