Joerg Reisenweber joerg at openmoko.org
Mon Apr 13 16:25:47 CEST 2009

Am Mo  13. April 2009 schrieb Steve Mosher:
> Dave Ball wrote:
> Save the new ID for a later device, targeted further in
> > the future.  But whether a 02.5 is sane would depend when it might enter 
> > MP.
>   Yup. Now you get why we had such a hard time pulling the trigger on it.
>   you have GTA02  shipping  an GTA02 -G that perform better and costs 
> less, which kills the 02 and an gta03 in flight which kills them both.
>   and then you had conflicts over engineering resource.

I really don't see the economics in doing a new engineering on schem, new 
layout, EVT etc pp, for a device we *hope* might be a _tiny_ improvement over 
what we are shipping now. I admittedly have no idea about marketing, but 

> > 
> > Now the buzz / bass problems are understood (fixed in new designs?) 
>   yes fixed

Bass on HS issue still unfixed even in GTA02A8. A 4u7 in place where there 
should be a 100u (47u absolute minimum) is a great improvement over the 1u we 
got now, but I wouldn't exactly call that fixed.

> The
> > Glamo seems to be the biggest weakness in gta02 - affecting UI 
> > responsiveness, 'wow' and software development of funky new stuff.
>    We'd
> > need a changed interface for the wifi (iirc), but from my non-expert 
> > position it seems reverting to the gta01 LCD interface should be low 
> > risk.  A simple gta02.5 should have a highly predictable schedule, and 
> > allow the new s/w stack to shine, with minimal software re-targeting 
> > needed.  If we kept RF & case designs from gta02 does that make 
> > certification easier?
>   Not really, probably just lowers the risk. Certs run  in parallel and 
> havent been critical path as far As I can recall.
> > 
> > gta03 / gta-future would then be (a little) further reaching, +p +e +cam 
> > + new ID +external SD card, and aimed at better defined 'consumers'.  
> > With more lead time, this is the one I see the community getting more 
> > involved in designing - identifying the target audience, appropriate 
> > parts, roughing out schematics, exploring limitations, rinse and repeat 
> > of the paper designs until OM are happy that it's buildable.

All this has been done ad nausea for GTA03. Just saying. (except the "define 
customers" part and the resulting rigid product specs)

> Interesting thought. Its probably a tall order to get the community up 
> to speed on the whole phone building business building process. I'd 
> start with a simple task.. GTA02 -G.. just to give people a taste of 
> what one is up against.

See my comment above. Doing the schem is easy like "i'll pass it over to you 
next afternoon" - especially if you got Mentor PADS for editing schem and 
layout (just because you *need* PADS to edit the project files, not to say 
PADS is anything very special or great). Proofreading schem, doing new 
layout, producing PCB and running EVT, just to find someone swapped the 100uF 
in application notes to a 1.00uF and you have no bass, then start all over 
again - incl new layout and PCB protos to find a space for a decent component 
instead of the faulty one... That's the real task.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
Url : http://lists.openmoko.org/pipermail/gta03/attachments/20090413/dd04560c/attachment.pgp 

More information about the Gta03 mailing list