I fired up my eTrex legend the other day, turned on WAAS, and it eventually
picked up satellite #46 (PRN-133). Everything was working great. The
receiver was receiving corrections and locked onto ranging on #46 (solid
bar). When I used it today, I saw #46 disappear and #51 (PRN-138) took over
in its place. It locked onto it and was receiving corrections, but because
the satellite is further west, the signal strength was not as good and some
of the regular satellites started to lose their corrections intermittently.
Since I'm located at 92°W longitude, PRN-133 at 98°W would be the highest
WAAS bird in the sky for me (and the east coast for that matter), so I would
think the receiver would choose that one first. I know my 12-channel
receiver can lock onto two WAAS satellites when there are less than 10 GPS
birds overhead, but with more than 10 it drops down to only one WAAS
satellite.

So my question is what determines which WAAS satellite the receiver picks?
Is the data from #51 judged to be somehow better than #46? I see PRN-133 is
flagged as NPA on the FAA's website, but does my non-aviation receiver even
know or care? It locked onto it fine so I'm somewhat mystified. I wish I
could force it to use 46 instead for better reception. Any insights?

Thanks,
Lee

I've never found a definitive explanation of Garmin WAAS
implementations.

There may be a hint here:

http://www.nstb.tc.faa.gov/RT_WaasSatelliteStatus.htm

Play the WAAS Satellite performance video and watch the status of the
three WAAS satellites.

--- CHAS

On Jan 1, 7:36*pm, "Lee" <l...@l.com> wrote:
> I fired up my eTrex legend the other day, turned on WAAS, and it eventually
> picked up satellite #46 (PRN-133). *Everything was working great. *The
> receiver was receiving corrections and locked onto ranging on #46 (solid
> bar). *When I used it today, I saw #46 disappear and #51 (PRN-138) tookover
> in its place. *It locked onto it and was receiving corrections, but because
> the satellite is further west, the signal strength was not as good and some
> of the regular satellites started to lose their corrections intermittently.
> Since I'm located at 92 W longitude, PRN-133 at 98 W would be the highest
> WAAS bird in the sky for me (and the east coast for that matter), so I would
> think the receiver would choose that one first. *I know my 12-channel
> receiver can lock onto two WAAS satellites when there are less than 10 GPS
> birds overhead, but with more than 10 it drops down to only one WAAS
> satellite.
>
> So my question is what determines which WAAS satellite the receiver picks?
> Is the data from #51 judged to be somehow better than #46? *I see PRN-133 is
> flagged as NPA on the FAA's website, but does my non-aviation receiver even
> know or care? *It locked onto it fine so I'm somewhat mystified. *I wish I
> could force it to use 46 instead for better reception. *Any insights?
>
> Thanks,
> Lee

Its possible that the WAAS satellite with worse geometry had a better
UDRE indicator, meaning its ranging (feature that allows WAAS
sattelites to provide GPS like time and positioning) was better. The
newest WAAS satellite (the one further east) is currently limited to a
50 meter UDRE, while other WAAS geos are allowed up to 7.5 meter UDRE.

Also WAAS corrections for non safety of life applications do not
require real time reception of the signal. As little as one full
minute of error free reception out of every 5 minutes should be enough
to keep you 100% up to date with corrections. Both IONO, clock and
ephemeris corrections are good for many minutes, because they are slow
changing. The clock correction include a velocity of change
indication. WAAS realtime reception is only important if you're flying
in the clouds close to the ground or walking on a minefield !

WAAS is really important when there are few satellites in view, 11
satellites in view doesn't make it useless, but it does reduce its
advantage quite a bit.

Marcelo Pacheco

On 2012-01-01 16:36 , Lee wrote:
> I fired up my eTrex legend the other day, turned on WAAS, and it
> eventually picked up satellite #46 (PRN-133). Everything was working
> great. The receiver was receiving corrections and locked onto ranging on
> #46 (solid bar). When I used it today, I saw #46 disappear and #51
> (PRN-138) took over in its place. It locked onto it and was receiving
> corrections, but because the satellite is further west, the signal
> strength was not as good and some of the regular satellites started to
> lose their corrections intermittently. Since I'm located at 92°W
> longitude, PRN-133 at 98°W would be the highest WAAS bird in the sky for
> me (and the east coast for that matter), so I would think the receiver
> would choose that one first. I know my 12-channel receiver can lock onto
> two WAAS satellites when there are less than 10 GPS birds overhead, but
> with more than 10 it drops down to only one WAAS satellite.
>
> So my question is what determines which WAAS satellite the receiver
> picks? Is the data from #51 judged to be somehow better than #46? I see
> PRN-133 is flagged as NPA on the FAA's website, but does my non-aviation
> receiver even know or care? It locked onto it fine so I'm somewhat
> mystified. I wish I could force it to use 46 instead for better
> reception. Any insights?

Garmin are quite mute on their algorithms, which I find frustrating.

The 'data' from the three WAAS sats is the same except for each
satellite's own ranging data. Hence your receiver should use whatever
_data_ it gets from any of the three N.A. WAAS satellites, but it should
not use the range from #133 (46) - for precision approaches. (WAAS is
foremost an aviation service). For ground pounding, it's safe, but we
can't assume what Garmin do with the range data on PRN-133.

I'm mystified by Garmin's use of WAAS. Here in eastern Canada
(Montreal) I even get EGNOS sometimes on my etrex 30 - which seems to
provide GLONASS corrections as well as GPS.

I do see #51 pop up on my photo tracker recording (Sirf III) more than
#46. (My cheap ($80) photo tracker is recording better, more accurate
tracks than the $300 etrex 30).

One thing is certain: at my latitude (45.5°) WAAS signals in the forest
are too weak to hang onto and I lose differential (both receivers).

I guess it's a line of sight issue when the receiver attempts to get a
WAAS satellite. First come first serve. In that light, when you turn
on your system, you might be able to use a corner of a building to mask
the sat you don't want and show the sat you want. Could be tedious and
more difficult in the southern areas of the northern Americas.

I don't recall the etrex 30 receiving more than 1 WAAS at a time, but
I'm pretty sure I've seen it receive both WAAS and EGNOS at the same time.

--
"We demand rigidly defined areas of doubt and uncertainty."
Douglas Adams - (Could have been a GPS engineer).

I went out with it again earlier today when the constellation was sparse
over my location. This time I was able to get both #46 and #51 locked at the
same time. Strangely enough, #51 actually had the stronger signal by about
20-25% over #46 even though it's a few degrees lower in the sky (perhaps it
has more transmit power). So perhaps that's why it was chosen. However, I
came back out later, and that time it was only using #46. I'm out in the
middle of a field with line of sight with the entire south sky, so shadowing
is not a factor.

I think I agree with the receiver evaluating the UDRE when choosing which
satellite to use. Signal strength might also be a factor. Perhaps when the
satellite geometry is poor, it will use #51 with the better UDRE to achieve
a better GDOP, but if the GDOP is already good then it will pick the closest
satellite (#46)?

I know DGPS corrections should be good for 5-10 minutes, but on my receiver
the D's on the satellite bars disappear if the WAAS satellite is out of
reception for only about a minute. I have no idea if it still uses the
corrections for some time after the D's are gone or not. Garmin's
algorithms are certainly mysterious.

On Jan 2, 8:52*pm, "Lee" <l...@l.com> wrote:
> I went out with it again earlier today when the constellation was sparse
> over my location. This time I was able to get both #46 and #51 locked at the
> same time. Strangely enough, #51 actually had the stronger signal by about
> 20-25% over #46 even though it's a few degrees lower in the sky (perhaps it
> has more transmit power). So perhaps that's why it was chosen. However, I
> came back out later, and that time it was only using #46. I'm out in the
> middle of a field with line of sight with the entire south sky, so shadowing
> is not a factor.
>
> I think I agree with the receiver evaluating the UDRE when choosing which
> satellite to use. Signal strength might also be a factor. Perhaps when the
> satellite geometry is poor, it will use #51 with the better UDRE to achieve
> a better GDOP, but if the GDOP is already good then it will pick the closest
> satellite (#46)?
>
> I know DGPS corrections should be good for 5-10 minutes, but on my receiver
> the D's on the satellite bars disappear if the WAAS satellite is out of
> reception for only about a minute. I have no idea if it still uses the
> corrections for some time after the D's are gone or not. *Garmin's
> algorithms are certainly mysterious.

Seems better to either:
1 - alternate between GEOs every two minutes and keep 11 channels for
GPS tracking. Or just enough to constantly refresh corrections from
each GEO and then move to the next
2 - track 10 GPS + 2 WAAS

Marcelo

On Jan 2, 2:52*pm, "Lee" <l...@l.com> wrote:
> Garmin's algorithms are certainly mysterious.

Garmin lists numerous changes to WAAS logic in their software revision
histories, but with no real details.

https://www8.garmin.com/support/down...ails.jsp?id=95

http://www8.garmin.com/support/downl...ails.jsp?id=26

"Improved WAAS operation when more than one WAAS satellite is visible.

Improve WAAS operation when transitioning from areas of ionospheric
data coverage.

Improved accuracy for WAAS EPE

Improved WAAS/EGNOS satellite selection algorithm to select the
satellite with the most beneficial corrections given the unit's
current position. A unit will not use a WAAS/EGNOS satellite if the
unit's current position is outside of a given WAAS/EGNOS satellite's
service volume.

Fixed potential memory corruption problem that could occur when
tracking more than one WAAS satellite simultaneously."

Dale DePriest speculated on how Garmin handles WAAS on the bottom
portion of this post, last updated about 4 years ago:

http://www.gpsinformation.org/dale/dgps.htm#waas

On 2012-01-02 19:43 , macpacheco wrote:
> On Jan 2, 8:52 pm, "Lee"<l...@l.com> wrote:
>> I went out with it again earlier today when the constellation was sparse
>> over my location. This time I was able to get both #46 and #51 locked at the
>> same time. Strangely enough, #51 actually had the stronger signal by about
>> 20-25% over #46 even though it's a few degrees lower in the sky (perhaps it
>> has more transmit power). So perhaps that's why it was chosen. However, I
>> came back out later, and that time it was only using #46. I'm out in the
>> middle of a field with line of sight with the entire south sky, so shadowing
>> is not a factor.
>>
>> I think I agree with the receiver evaluating the UDRE when choosing which
>> satellite to use. Signal strength might also be a factor. Perhaps when the
>> satellite geometry is poor, it will use #51 with the better UDRE to achieve
>> a better GDOP, but if the GDOP is already good then it will pick the closest
>> satellite (#46)?
>>
>> I know DGPS corrections should be good for 5-10 minutes, but on my receiver
>> the D's on the satellite bars disappear if the WAAS satellite is out of
>> reception for only about a minute. I have no idea if it still uses the
>> corrections for some time after the D's are gone or not. Garmin's
>> algorithms are certainly mysterious.
>
> Seems better to either:
> 1 - alternate between GEOs every two minutes and keep 11 channels for
> GPS tracking. Or just enough to constantly refresh corrections from
> each GEO and then move to the next
> 2 - track 10 GPS + 2 WAAS

12 channel receivers are borderline obsolete in the age of SBAS.

IAC, while we sometimes see upwards of 11 or more GPS satellites,
permitting the tracking of 2 geo sats, the better (safety) algorithm
would be to always attempt to track 2 SBAS's and up to 10 GPS'.

When you have 10 GPS' the horizontal error will be minimized even when
discarding the very lowest satellites even if above the mask angle.

For ground pounding however, w/o precision requirements, a single SBAS
is normally enough.

--
"We demand rigidly defined areas of doubt and uncertainty."
Douglas Adams - (Could have been a GPS engineer).

On 2012-01-02 17:52 , Lee wrote:
> I went out with it again earlier today when the constellation was sparse
> over my location. This time I was able to get both #46 and #51 locked at

How sparse? How well were the sats distributed?

> the same time. Strangely enough, #51 actually had the stronger signal by
> about 20-25% over #46 even though it's a few degrees lower in the sky
> (perhaps it has more transmit power). So perhaps that's why it was
> chosen. However, I came back out later, and that time it was only using
> #46. I'm out in the middle of a field with line of sight with the entire
> south sky, so shadowing is not a factor.
>
> I think I agree with the receiver evaluating the UDRE when choosing
> which satellite to use. Signal strength might also be a factor. Perhaps
> when the satellite geometry is poor, it will use #51 with the better
> UDRE to achieve a better GDOP, but if the GDOP is already good then it
> will pick the closest satellite (#46)?
>
> I know DGPS corrections should be good for 5-10 minutes, but on my
> receiver the D's on the satellite bars disappear if the WAAS satellite
> is out of reception for only about a minute. I have no idea if it still
> uses the corrections for some time after the D's are gone or not.
> Garmin's algorithms are certainly mysterious.

Don't forget that the primary benefit of WAAS is not the range from that
satellite, but the data regarding the other satellites. That's where
most of the increased precision comes from. Having an additional
satellite for range (the WAAS sat), even with a sparse constellation, is
not improving your horizontal accuracy very much unless you are pretty
far north.

While Garmin' algorithm's are mysterious, I would bet that when the
"D"'s disappear, the differential data for that sat has been discarded.

--
"We demand rigidly defined areas of doubt and uncertainty."
Douglas Adams - (Could have been a GPS engineer).

On 2012-01-02 15:02 , Alan Browne wrote:

> I'm mystified by Garmin's use of WAAS. Here in eastern Canada (Montreal)
> I even get EGNOS sometimes on my etrex 30 - which seems to provide
> GLONASS corrections as well as GPS.

That one's solved. I didn't realize that EGNOS covers parts of North
America:

http://egnos-portal.gsa.europa.eu/di...rmation/status

--
"We demand rigidly defined areas of doubt and uncertainty."
Douglas Adams - (Could have been a GPS engineer).