[time-nuts] GPS altitude somewhat wrong?

[Gary E. Miller]

Yo Tom!

On Wed, 8 Jun 2016 19:26:21 -0700
"Tom Van Baak" <tvb at LeapSecond.com> wrote:

> From one NW GPS farm to another... I'm willing to help you debug
> yours.

Nothing to debug.  I know if I move the antenna I can do better, but
it does what I need, and it is near the server that needs it. ntpd
caculates precision -22 and the jitter is usually less than 0.5 uSec.
More than good enough for my purposes.  It makes a dandy Stratum 1 
chimer.

And 16 meters is not good at all.  That is 52 feet!  Over just two
hours.  I bet that gets 2x or more worse over 24 hours.  If I cared
about location accuracy I would throw away that GPS. Any modern GPS
should do way better than that.

> > A Garmin 18x reports:
> >    Altitude Err:    +/- 264 ft     
> 
> Something is terribly wrong with your setup. The Garmin 18x is much,
> much better than this. I know because the 18x was one of the GPS
> receivers I brought along on a recent mobile clock experiment.

Yes, with good antenna placement it can be better, but the 18x also
degrades much quicker than newer GPS when the antenna placement is places
badly.

> in Tucson: http://leapsecond.com/great2016a/2016a-garmin-18x-2.gif

what year is yours?  Looking at mine, I see it is actually an 18, not an
18x, so a 12 year old design .  Worse than the 18x.

> With clear sky view, the peak to peak is under +/- 8 m, and the
> (1-sigma) standard deviation is 3 m. Even at the hotel lobby, with
> obstructed sky view, the (1-sigma) standard deviation stayed under 7
> m. Your 18x number, +/- 264ft (+/- 80 m), is 10x to 25x worse than
> this. It doesn't feel right.

My antenna is between two 2 story houses at ground level against a tall
hill.  It is lucky to get much of a signal at all.  I know it would work
better in aother location, but it needs to be there next to my main
servers.

> Off-list, can you send me a day of NMEA from your 18x? Not gpsd
> output; but the raw serial ascii data from the receiver. I'd like to
> get to the bottom of this. We'll all learn something.

I can get you pseudo NMEA, the Garmins work much better in binary mode.
When they work in binary mode. :-)

I'll start to grab that now.  I'll have a 12 hour scatter plot in
the morning.

While you are waiting check out the attached scatter plot.  Now THAT
is a good $25 GPS!  Beats the heck outta any Garmin.  CEP(95) of 1.5
meters over 1,000 seconds.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
	gem at rellim.com  Tel:+1 541 382 8588
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screenshot_2016-04-19_17-18-24.png
Type: image/png
Size: 82984 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20160608/1994a01d/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20160608/1994a01d/attachment.sig>