[time-nuts] how good an oscillator do you need for a GPS simulator

Jim Lux jimlux at earthlink.net
Fri Dec 16 14:47:52 UTC 2011


On 12/15/11 10:25 PM, Chris Albertson wrote:
> For testing, I'd assume the gps simulator only needs to be good enough
> that the receiver will detect the signal.   There is some Doppler
> shift so the receiver must have to look over a wider range of
> frequencies so if the simulator was inside that range it could work.
>   Light travels at about one foot per nanosecond.   so your simulator
> should need to know the time to within a few tens of nanoseconds.
> Receivers can deal with not-perfect signal.  Multipath and refraction
> are common.
>

Not just functional test, but to verify the "added noise" from the 
receiver.

TO just see if it can acquire and track, pretty crummy would work, 
because it's just like the horrible signals it's seeing "off the air"


> You GPS simulator would likely have a GPS receiver inside of it and
> sync to a real GPS.

Not necessarily.  You might be testing in a screen room with no external 
signals available.

Clearly, one can go out and spend 500k on a nice Spirent, but I'm 
looking at what is the few hundred dollar solution.  A reasonably quiet 
oscillator driving an FPGA or playing back bits from RAM



> On Thu, Dec 15, 2011 at 9:37 PM, Jim Lux<jimlux at earthlink.net>  wrote:
>> Say you want a quik n easy n cheap GPS simulator to test a GPS timing
>> receiver.  How good does the oscillator (presumably some nice multiple of
>> the chip rate) have to be?





More information about the Time-nuts_lists.febo.com mailing list