Reason for GPS problems found!

Robert Horton openmoko at roberthorton.net
Tue Jul 15 23:50:50 CEST 2008


> -----Original Message-----
> From: community-bounces at lists.openmoko.org [mailto:community-
> bounces at lists.openmoko.org] On Behalf Of Joerg Reisenweber
> Sent: Tuesday, July 15, 2008 4:29 PM
> To: community at lists.openmoko.org
> Cc: thomasg
> Subject: Re: Reason for GPS problems found!
> 
> Am Di  15. Juli 2008 schrieb thomasg:
> > Hi ppl,
> >
> > I write this to community, not to devel or owners because everyone
> > should
> > know:
> > *sbeh*, one of the people in #neo1973-germany IRC-channel found the
> > reason for the GPS problems.
> > The problem only occurs if a SD card is set in. Doesn't matter if
> it's
> > mounted or in use, it just has to sit in the socket.
> > The TTFF went from no fix at all to TTFF 120 seconds indoor(!!!), and
> > about 40 seconds outdoor.
> > Two other people could verify this with about the same results.
> > We'll do more tests later, but for now we surely know what's causing
> > the problem (and it seems to be a EMC problem).
> >
> > First results show at the same devices, even outdoor, that there is
> no
> > fix in over 400 seconds with SD card, the signal seems to be at least
> > 10 to 20 dB worse (so bad, that most satellites don't even appear).
> >
> > Testresults from other people appreciated.
> 
> 
> 
> So we can assume this is pretty confirmed by quite some reports now.
> Thanks to community for the awesome help on this!
> Special thanks to *sbeh* for finding this. Please mail me directly ;-
> )!!
> 
> We did quite some tests on this as fast as we could (and I never
> assumed we could be *that* fast), and it seems we found the root cause
> of the problem.
> ALL THIS IS PRELIMINARY INFO! Due to further tests to confirm.
> Just to let you all know we're on it, and please stop thinking about
> shielding the SDcard with tin foil or relocating internal gps-antenna.
> It won't probably help.
> 
> The good news: *IF* all pans out, there's (or soon will be) a new
> kernel at Andy branch that stops SD-card clock when SDcard is idle. We
> hope this will almost cure the problem, at least reduce it to sth like
> "you can't GPS while watching video from SD" or the like (hope you can
> cope with that ;).
> We're about to verify a hw-fix so you could even watch video and still
> have GPS positioning during that.
> 
> 
> Please stay tuned and be patient, we'll come up with news *really
> soon*.
> And *please* don't try weird reworks and maybe break your FR by doing
> so
> ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-) ;-)
> 
> cheers
> jOERG


Thanks for all this fast testing and great work. I was worried when there
was silence. I hope for both a hardware and software fix, because if the
solution is just a patch to turn off the SD while getting a fix, it may work
but I feel that I have inferior hardware...even if it's on a psychological
level only. With truly fixing the root cause of the interference via
hardware then I know that even if the GPS signal gets faint in a condo
canyon or tunnel the SD card won't hinder GPS operation.

That way when I take the FR into the Houston tunnels [1] I won't be cursing
the SD card when it's really the impossible chance of getting a signal in
the basement of a skyscraper of downtown Houston. ;-)

[1] http://en.wikipedia.org/wiki/Houston_Downtown_Tunnel_System


-Robert






More information about the community mailing list