GPS time synch

[Ian Stirling]

Perry E. Metzger wrote:
> Ian Stirling <openmoko at mauve.plus.com> writes:
>> Based on ballpark estimates from other similar devices, it's probably
>> around 70mW.
>>
>> This will really hurt if you keep it on all the time, fortunately, for
>> most applications, you probably don't need to.
>> If it works at 2s on, 30s off, as seems likely, this alone drains
>> 1/30th of the battery per day, which isn't bad. Especially as you can
>> probably back off a little if the phone has been stationary for
>> several minutes.
> 
> Does GPS work well in such a mode? I seem to remember from various
> handheld units that it can take quite a while (minutes) to sync up to
> a given satellite. (For all I know, you *can* operate this way, I'm
> completely clueless on the subject.)

Several cases.
'cold start' - this can take up to around several minutes to get an 
accurate position.
The only time this happens is when the GPS has not downloaded an almanac 
in 6 weeks (either directly from the satellites (requiring being on for 
13 minutes) or from the internet. Or hasn't been on for 30s every 3 or 4 
hours since it last did get an almanac.

If the neo has a GPS signal when it's charging, this (several minute 
locks) will never happen.

To get lock in a couple of seconds requires a pretty good idea of your 
position, the time, and maybe even being able to tell the hardware chip 
the estimated relative speeds of the satellites.

Again, if you're turning on and off the GPS every minute or so, and 
you're not expecting it to cope with continent sized jumps, it's not a 
big deal - especially if the hardware can be fed estimates, rather than 
trying to figure it out on its own, which it can, but it may  be slower.