[time-nuts] Accuracy/drift of Garmin GPS 16 HVS 1 PPS output under invalid fix conditions...

[Tom Van Baak]

> Just a basic question for an application with low accuracy 
> requirements. I only need an accuracy of +/- 1 ms on the 1 PPS signal.
> 
> Does anyone have any data on the behaviour of a Garmin GPS 16 HVS (or be 
> able to point me in the right direction to find such data) w.r.t. the 1 
> PPS output when a valid fix is not available.
> 
> I need this to determine whether the 1 PPS hold-over accuracy is 
> sufficient to maintain the 1 PPS within +/- 1 ms of GPS time during a 
> loss of GPS lock of, say, 15 minutes.
...
> Steve Olney

Hi Steve,

Ok, this was an interesting one because what you're asking is so different from how GPS/1PPS is typically used by anyone.

As you know these Garmin OEM receivers have a configuration mode where it will still continue to output the 1PPS pulse even though reception is lost and even though the receiver knows the 1PPS is in error. What you're asking is how large is the error. It's a fair question.

I've used the Garmin model 18 a lot over the years, and I also have a Garmin model 16 like yours. So I tried to duplicate what you're asking on both receivers and the results are quite interesting. Normally we time nuts only measure GPS receivers they're outputting a clean 1PPS. Most receivers wisely disable the 1PPS when reception is lost. But by using SNSRXCFG.exe you can configure these Garmins output the 1PPS even when its wrong. The "auto-off disable" mode allows you to see how dirty the 1PPS becomes when reception is lost.

The short answer is I would trust my model 18 to stay within 1 ms over 15 minutes, but I would not have similar trust in the model 16. Here are lots of plots for you:

http://leapsecond.com/pages/garmin-gps/

/tvb