Revolution Reading GPS Data

[Ben Rubinstein]

On 4/3/08 10:45, Graham Samuel wrote:
> FWIW, the GPS device I described is marketed by a European sports 
> retailer, Decathlon, as KeyMaze 300 GH-601. I've found out via the SiRF 
> web site (the device uses a SiRFstarIII processor) that it is actually a 
> rebranded GlobalSat GH-601 - this is GlobalSat of Taiwan. On the SiRF 
> site this is further described at
> 
> http://www.sirf.com/success_10.html
> 
> Annoyingly the device is not described on GlobalSat's own web site (a 
> site search reveals nothing), but strangely their FAQ pages provide a 
> little info about downloading firmware for it etc. I suppose this means 
> that either the thing is obsolete or else my retailer Decathlon has made 
> an exclusive deal to market it - they are certainly selling it quite 
> vigorously. I tried emailing GlobalSat to see if they have any more 
> information available, but they didn't reply. I guess they don't deal 
> with consumers.
> 
> The device comes with a bit of PC software which is just about adequate, 
> but I would not say that it was well-written or complete - its UI is 
> crude and it doesn't even help to file all the information which it 
> extracts from the device, and indeed it may be throwing away information 
> (such as timestamps) which is just not visible to the user.
> 
> It looks to me as if my wish to create better software for it (on a Mac 
> primarily) is pretty much a dead end - although if I can get hold of a 
> serial-to-USB converter I might be able to experiment a bit.

...not wishing to stop you from getting back to your day job, but you might 
not be completely at a dead end.

AFAICT, almost all GPS devices are using SiRFStar II or III as the actual GPS 
chipset.  (And FWIW a very large number of them are using Globalsat technology 
wrapped round that, regardless of the name on the box.)  And at root these 
things output serial; so they are glued onto (I don't know much about 
electronics, can you tell?) a serial-USB chip in order to give them a USB 
interface.

There seemed when I was working in this area to be two main sources of the 
latter: FTDI and ??Prolific?.  Although the devices I was working with were 
Windows only, I was able to find Mac drivers for both of these things with a 
bit of googling on the net (because other products are using the same 
chipsets, and are marketed with Mac software).  Once I got these installed I 
could simply open a "port" called /dev/cu.usbserial, and found that I was 
talking to the GPS device as if it was a serial device.  So in fact Rev was 
dealing with a serial device; and the actual SiRF chipset was a serial device; 
there were just a few layers of bridging over USB going on between the two.

So I reckon that there's a fair chance that you could establish communication 
with the device (though mind you I did this a couple of years ago, on a 
PowerPC - I've not checked whether Intel versions of these drivers are now 
available).

However: I was dealing with basic GPS devices, no memory, interface, or 
anything: so was essentially just trying to talking to the SiRF chip to get 
the current GPS data in realtime.  It sounds as though you might have a device 
that actually does stuff to record your trace etc when you're out and about, 
and then you plug it in later and extract info.  In that case you're probably 
talking to some other piece of hardware, which has done the communication with 
the SiRF, stored the results, and so on; in that case the device you're 
actually talking to may not "talk serial", and in any case you'd have to 
reverse-engineer the protocol, which would presumably be proprietary (unlike 
the SiRF, which talks NMEA).  This may not be that impossible (judging by the 
macam project, which has succesfully created Mac support for hundreds of 
digital cameras).  I believe that there is a handy utility available that lets 
you eavesdrop on USB traffic, which might be the place to start...

good luck!

- Ben