[time-nuts] GPS 18 behavior

Jim Lux jimlux at earthlink.net
Sun Jul 21 13:55:56 EDT 2013


On 7/21/13 9:28 AM, David J Taylor wrote:
> From: Jim Lux
> []
> Oh, I didn't actually think it would be 10% off.. more like a few ppm,
> depending on temperature.  I was wondering more what would happen if you
> were indoors for a couple days, then went back out, what the box does.
> realistically, I don't expect that it smoothly brings it back into line,
> rather, I would expect it to "jump" to the correct time and timing.
>
> I think, realistically, I should be looking at the actual sentences
> coming out and figuring out whether our "GPS synchronized" state has
> changed, and maybe deal with it appropriately.
> =====================
>
> Jim,
>
> My experience is limited, but I recall that the PPS stopped soon after
> the signal disappeared, and would only re-appear when the unit had
> sufficient lock on the GPS.  There is indeed a valid/not-valid bit in
> the output stream (perhaps referring to position, it's in the manual).


when I'm in a GPS denied environment, it's not just because we're 
indoors, it's because we're somewhere that GPS isn't available, so what 
I'm really doing is providing a sort of flywheel to keep my little 
modules synced with each other.  I don't need super accuracy in an 
absolute sense: I just need to tag the data all with the same time tag 
within a few seconds.

I have a GPS in the system, so that was convenient.


>
> My GPS 18x LVC is noticeably more sensitive than the GPS 18 LVC, and
> will work almost all of the time indoors.  <snip>

> You might like to go for an 18x and put it in view of the sky, if
> possible, or even use a module like:
>
>   http://www.adafruit.com/products/746
>

Mostly, I'm trying to use what's already in the system.  I have the 
GPS-18 already there feeding a PC with position and time (when 
available), and as long as it has the 1pps, I figured I could use that 
to sync the modules with a hardline (that talk to the host PC via USB).

Right now, I sync them (without GPS) by having one be a master and send 
a pulse that the others receive.  They're little "teensy3" modules (like 
a fast Arduino with more memory and peripherals), so I just have one of 
the GPIO pins wired together.

Somewhere down the road, I'm going to be changing the system 
architecture so that the modules aren't all in the same physical box and 
a wire for sync isn't practical.  So the idea is to put a GPS with each 
module, and assuming they all see the GPS at the same time, they each 
sync with the 1pps.

If I can figure out how get the "sync to garmin, but have flywheel" work 
in the present system, then I can  start to get that piece of the future 
system done now.  Or, at least, go through the first half dozen iterations.

In a very much down the road scenario, I'll be looking at trying to 
synchronize the modules to 100 microseconds AND try to make the RF 
section stay on the same frequency within a fraction of a Hz (at 
3-30GHz).  *that* is a substantially more challenging chore.  Basically 
implementing a GPSDO good to 1E-11 over 30 seconds in each module.  And 
small, and battery powered, and cheap, of course <grin>

For today, though, sync to 1 millisecond among the physically connected 
modules is plenty good.



More information about the time-nuts mailing list