| |
| ProfessionGPS |
|
|
| POIenFrance |
|
|
| Visiteurs |
Actuellement en ligne: 425
Dont membres: 3
+ d'infos...
|
|
|
| GPS Chipsets and Receivers Compared |
2. Overview of Specs and Observations - SBAS (WAAS/EGNOS)
In keeping with the approach of the site since it was started in 2002, this article will focus on sharing the field testing experience. Since the only way I know how to this is by testing GPS receivers, I set out to select a representative GPS receiver for each chipset. Some might argue that performance will depend on the design of the receiver but I've found in my daily test runs of the numerous SiRF receivers on the market that performance is quite consistent between receivers using the same chipset. Exceptions, if any, will be noted in the page dedicated to each chipset.
Let's get started by looking at what the chipset designers have to say about their chipsets on their websites (see below for the links), after all without them we wouldn't be discussing the finer points of GPS performance! This is what the first part of the table does by compiling bits and pieces collected on various sites. I spent more time than I expected putting this information together and I'm not sure many conclusions can be drawn from them, but at least this can serve as a reference in the future. What does jump out are the ultra low power rating of the NemeriX NJ-1030 chipset, the SiRF XT2 and SS3 sensitivity ratings (more numbers and discussions here) and of course the massive correlation power of the SiRFstarIII chipset. Without going into too much detail, correlation is the action of analyzing the signal received from the satellites and making sense out of it to compute a position. Sounds simple enough, but the signals are very weak (sent from 20,000 kms above our heads) and generally very degraded by our immediate environment, so the more correlation power you have the better your chances are of getting a "fix".
Concerning SBAS or WAAS as it is known in the US and EGNOS in Europe. As can be seen from the table, chipset designers don't seem to be too "anxious" to implement SBAS (Update - available on SiRFIII since v3.1 and Nemerix v4.0), which may come as a shock to some, but is easily explained by the fact that since Selective Availibility (SA) was removed in May 2000, the residual GPS error of approximately 10 meters is mostly due to GPS signal deflection that the SBAS system can only marginally correct (I have never been able to measure any improvement from EGNOS but others have seen 1/2 meters with WAAS). On the other hand SBAS is crucial for critical use such as flying a plane as it allows for near real time GPS signal integrity checking. Also check this discussion in the technical forums.
...the second part of the table shows the "mapping" of the GPS receiver to the GPS chipset. I obviously had a lot of SiRF II/LP and XT2 based receivers to chose from so I picked those I use most. The battery testing does show that NemeriX has the lead there although not with the ratio shown in the specs, but the draw of the antenna and Bluetooth modules need to be factored in. The average perceived TTFF (Time To First Fix) and Sensitivity are just that, a summary of the subjective impressions I collect by using and testing these receivers on a daily basis. The following pages will focus on objective measurements, and we'll see if these impressions need to be mitigated.
One quick word about SiRFstarIII, the massive correlation power indicated in the specs does seem to play out on the field with GPS performance becoming more consistent, TTFF is less dependent on the environment and once you get a lock you rarely lose it, even indoors. With SiRFIII, GPS requires less "work", for instance, you no longer need to choose between waiting 2 minutes when you leave a parking garage in the morning to get a fix or risk driving around for 10 minutes without one, now you'll generally get a fix in 45 seconds in both situations .
On to the tests...
If you have questions or comments, you can use this discussion thread in the forums
|
|
|
|
|
|
del.icio.us
