More on SiRFStarIII.
I've now spent a good deal of time working with three SiRFStarIII-based GPS modules, two of USGlobalSat's BT-359W, and an EM-406 module. http://www.usglobalsat.com/item.asp?itemid=134 http://www.usglobalsat.com/item.asp?itemid=131
All three have shown difficulty capturing WAAS corrections from PRN122, even though the indicated signal is strong and the satellite is 46 degrees above the horizon here. There are few comments online about this difficulty, but others have seen it, too. I'm waiting for a response from USGlobalSat.
BTW, DGPS status is indicated in the sixth field of $GPGGA.
Tom
GPS Module Insight
GPS Timing
I, too, am working with the EM-406 GPS module, from SparkFun.
Nice unit, good sensitivity for a patch antenna, rapid signal lock, works indoor, etc.
Timing, however, is everything.
$GPRMC packets arrive once per second. (Standard NMEA Format).
I used this for the Master Timing Reference for a program, with numerous Tasks running in the background. I kept getting a periodic data error.
Turns out that the RMC packet is sent once per second, however the interval between its transmission is not fixed at precisely 1 second intervals. Based on the default packet transmissions, and the variable length GSV packets, the intra-RMC timing is:
1.0 1.0 1.0 1.3 0.7 (Seconds) repeat.
Live and Learn, and back to the drawing board...
JC
Nice unit, good sensitivity for a patch antenna, rapid signal lock, works indoor, etc.
Timing, however, is everything.
$GPRMC packets arrive once per second. (Standard NMEA Format).
I used this for the Master Timing Reference for a program, with numerous Tasks running in the background. I kept getting a periodic data error.
Turns out that the RMC packet is sent once per second, however the interval between its transmission is not fixed at precisely 1 second intervals. Based on the default packet transmissions, and the variable length GSV packets, the intra-RMC timing is:
1.0 1.0 1.0 1.3 0.7 (Seconds) repeat.
Live and Learn, and back to the drawing board...
JC
NMEA output is not usually well synchronized with 1pps since the serial output processing is a background firmware function in most GPSR processors, of low priority. A TTL 1uS-accurate 1pps is available on the EM-406, though - if you have a steady soldering hand - on TP10, and pin 6 on the connector is free.
Tom
GPS Module Insight
SiRFStarIII: WAAS than I thought.
I just concluded a simultaneous comparison of the SBAS channel function
of a new SiRFStarIII chipset in a GlobalSat BT-359W (said to be
WAAS-enabled) and a SiRFStarII chipset in an older BU-303 (WAAS-capable
but not enabled by default).
From southwest Florida, the new BT-359W recognizes a signal from
geostationary PRN 122 (Inmarsat 4F2), but it does not decode the
corrections despite an apparently excellent signal. That is the only
SBAS signal that this receiver indicates.
Conversely, once enabled the BU-303 gets EGNOS corrections from PRN 120
(Inmarsat 4F2), and WAAS corrections from PRN 122 (3F4), PRN 135 (Galaxy
15) and PRN 138 (Anik F1R). It also points to but can't see PRN 126 and
PRN 134.
Oddly, the two receivers think that PRN 122 is a different satellite.
The BT-359W points to 4F2 at 53W, while the BU-303 points to 3F4 at
142W. If the BU-303 eventually recognizes 4F2, it would have access to
five WAAS/EGNOS sources from here.
Still, even without the WAAS differential corrections, the SiRFStarIII
chipset well outperforms its predecessor. I want what is promised,
though, from these WAAS-enabled units, and they do not apparently
deliver it.
I have two iTrek M5 Bluetooth GPS receivers coming; these are
Nemerix-based receivers that are reported to perform close to the
SiRFStarIII and are said have no difficulty with PSAS.
Tom
I just concluded a simultaneous comparison of the SBAS channel function
of a new SiRFStarIII chipset in a GlobalSat BT-359W (said to be
WAAS-enabled) and a SiRFStarII chipset in an older BU-303 (WAAS-capable
but not enabled by default).
From southwest Florida, the new BT-359W recognizes a signal from
geostationary PRN 122 (Inmarsat 4F2), but it does not decode the
corrections despite an apparently excellent signal. That is the only
SBAS signal that this receiver indicates.
Conversely, once enabled the BU-303 gets EGNOS corrections from PRN 120
(Inmarsat 4F2), and WAAS corrections from PRN 122 (3F4), PRN 135 (Galaxy
15) and PRN 138 (Anik F1R). It also points to but can't see PRN 126 and
PRN 134.
Oddly, the two receivers think that PRN 122 is a different satellite.
The BT-359W points to 4F2 at 53W, while the BU-303 points to 3F4 at
142W. If the BU-303 eventually recognizes 4F2, it would have access to
five WAAS/EGNOS sources from here.
Still, even without the WAAS differential corrections, the SiRFStarIII
chipset well outperforms its predecessor. I want what is promised,
though, from these WAAS-enabled units, and they do not apparently
deliver it.
I have two iTrek M5 Bluetooth GPS receivers coming; these are
Nemerix-based receivers that are reported to perform close to the
SiRFStarIII and are said have no difficulty with PSAS.
Tom
Tom
FWIW, SiRFStar-III followup:
US GlobalSat accepted two trouble reports for no-WAAS on a pair of BT-359Ws, said to be WAAS-enabled but which never find WAAS satellites that are in easy view. An EM-406 suffers the same difficulty, and _every_ online report is similar.
With no communication whatsoever except my trouble report, they issued an RMA; I sent the unit to them. 10 days later I received a new identical receiver; same hardware, same firmware, same problem. GlobalSat is delaying a fix to the problem as long as they can, it appears. Sad.
For the time being, if you are a SiRFStar-III user, don't expect WAAS (or any SBAS corrections) from this chipset; look elsewhere.
US GlobalSat accepted two trouble reports for no-WAAS on a pair of BT-359Ws, said to be WAAS-enabled but which never find WAAS satellites that are in easy view. An EM-406 suffers the same difficulty, and _every_ online report is similar.
With no communication whatsoever except my trouble report, they issued an RMA; I sent the unit to them. 10 days later I received a new identical receiver; same hardware, same firmware, same problem. GlobalSat is delaying a fix to the problem as long as they can, it appears. Sad.
For the time being, if you are a SiRFStar-III user, don't expect WAAS (or any SBAS corrections) from this chipset; look elsewhere.
Tom
some time ago, I fiddled with receiving correction data via the Internet. Free from some university. I recall that if the reference receiver is within 500mi of you, the corrections are generally good enough for about 15 minutes' use, i.e., for IDOP dominated error compoents, assuming good GDOP). But GPS receivers imply mobility and Internet connectivity while mobile is of course problematic - limited to say Verizon EV-DO for good coverage.
If your needs are data logging of location, and you can stand delayed correction, there is software to do post-processing correction, given you take the raw pseudo-ranges from the receiver and time/stamp/log them.
If your needs are data logging of location, and you can stand delayed correction, there is software to do post-processing correction, given you take the raw pseudo-ranges from the receiver and time/stamp/log them.
In fact, RTCM SC-104 DGPS corrections can be carried by any path, and there are many available including IP, broadcast FM subcarrier, pager, LF beacon and satellite, both free and commercial. WAAS is just a handy method that uses the same frequency as GPS, but from geostationary satellites.
I'll probably eventually use LF beacons for this marine project, but I'm trying to use WAAS for convenience while in development. Where this system will be useful there are usually several good LF beacons that offer local corrections. DGPS boaters here are well accustomed to 2-meter accuracy routinely. WAAS can't do as well, theoretically, since it only uses about 20 ground stations as references.
I am most disappointed with the guaranteed SiRFStar-III chipset WAAS failure (despite the company's implication that WAAS will work by selling "WAAS-enabled" products), but the chipset makes a superb GPS receiver in every other regard. Bummer.
I'll probably eventually use LF beacons for this marine project, but I'm trying to use WAAS for convenience while in development. Where this system will be useful there are usually several good LF beacons that offer local corrections. DGPS boaters here are well accustomed to 2-meter accuracy routinely. WAAS can't do as well, theoretically, since it only uses about 20 ground stations as references.
I am most disappointed with the guaranteed SiRFStar-III chipset WAAS failure (despite the company's implication that WAAS will work by selling "WAAS-enabled" products), but the chipset makes a superb GPS receiver in every other regard. Bummer.
Tom
All I remember about WAAS is that the pacific and atlantic satellites are parked way out over the oceans for ships at sea INMARSAT, so GPS receivers have a very low grazing angle on the US coasts. And inland it doesn't take much terrain to obscure the satellites. A mid-No. America satellite was planned.
GPS Module Insight
> ... A mid-No. America satellite was planned...
Yes, that's been fixed. An old Inmarsat (PRN122) was moved from 53W to
142W and a new Inmarsat is now at 53W, although it does not carry
SBAS/WAAS. Still, from south Florida there are four visible WAAS
satellites and one visible EGNOS satellite; western US WAAS coverage is
even better.
Yes, that's been fixed. An old Inmarsat (PRN122) was moved from 53W to
142W and a new Inmarsat is now at 53W, although it does not carry
SBAS/WAAS. Still, from south Florida there are four visible WAAS
satellites and one visible EGNOS satellite; western US WAAS coverage is
even better.
Tom
I finally got tired of being put off by USGlobalSat and opened a thread on the Belgian GPSPassion.com site, which is well-read. To shorten a frustrating story, someone at GlobalSat finally acknowledged, privately, that they were aware of my no-WAAS BT-359W complaints, but his only remedy at the time was "Hmmm".
Increasingly furious, I found a phone number and pounded the keypad. After assurance that - "just last week" - they'd received new stock with v3.2.2 firmware in them, I sent the two v3.1.1 units off and got two BT-359C replacements yesterday. In fact, finally, WAAS works on the SiRFStar-III chipset.
The flash date on the unit, though, is October, so the cynic in me figures they avoided shipping the new firmware until they'd lowered stock of the previous version (they cannot, I'm told, reflash the firmware without sending them to Taiwan).
Most customers surely don't need WAAS so the earlier version is probably fine for them, but the box and advertising still says "WAAS-enabled default unit" - and, indeed, that is true but misleading; a faulty process, that cannot work, was enabled. That's just shy of fraud.
Still, I finally have WAAS-working SiRFStar-III-based receivers.
Increasingly furious, I found a phone number and pounded the keypad. After assurance that - "just last week" - they'd received new stock with v3.2.2 firmware in them, I sent the two v3.1.1 units off and got two BT-359C replacements yesterday. In fact, finally, WAAS works on the SiRFStar-III chipset.
The flash date on the unit, though, is October, so the cynic in me figures they avoided shipping the new firmware until they'd lowered stock of the previous version (they cannot, I'm told, reflash the firmware without sending them to Taiwan).
Most customers surely don't need WAAS so the earlier version is probably fine for them, but the box and advertising still says "WAAS-enabled default unit" - and, indeed, that is true but misleading; a faulty process, that cannot work, was enabled. That's just shy of fraud.
Still, I finally have WAAS-working SiRFStar-III-based receivers.
Tom