[time-nuts] Quality of timing mode GPS vs survey accuracy

[Bob kb8tq]

Hi

> On May 8, 2020, at 5:12 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> 
> 
> Is there any published data on what happens to the quality of time if the 
> survey is off by xxx meters?

Speed of light plus distance off “correct” plus constellation geometry are what 
are usually tossed into examples. Turning that into an exact error number for
this receiver design at time = x …. never seen it taken that far. You could do it 
yourself for a hypothetical processing approach. 

> 
> Do all the GPS receivers use the same coordinate system?

GPS uses a coordinate system, receivers translate this into another system to display
the results. Many can be set to a variety of coordinate systems.  Unless there are
bug in the firmware, when set to the same system they should be talking the same
way. Mr Google will give you a long list of hits going into this for each GNSS system. 

>  If I take a survey 
> with several GPS receivers will they all get the same answer?

Nope. 

If you take a survey with the *same* receiver, there will be some level of “this pass
does not agree with that pass”. 

How big the errors are depends a lot  on:

1) what sort of receiver you have (single band / multi band / design era / make / model)
2) what sort of antenna you have (fancy geodetic / simple pole mount / mag mount …)
3) how well you have things set up (sky view ….)
4) how bad your local situation is ( multi-path, jammers, RFI / EMI ….)
5) how long the survey is ( minutes / hours / days / weeks / month ….) 
6) what the conditions are (space weather / local weather / GNSS system issues …)
7 )what sort of post processing you do ( antenna corrections …. IGS data sets used ….)

(and on and on and on ….)

For timing, you may not care about errors dimensioned in mm, but there are people
who do. Yes, this is a basic issue with wide open questions :)

Bob

> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.